Notes from a Guangzhou
Garment Manufacturer

MOQ 30 costs more per piece than 500. Any small MOQ factory will admit that much on the first call. The part nobody puts in the listing: 30 pieces is still the cheaper decision for a first run, and it isn't close.

What does 30 actually cost next to 500? Per piece, more — we add a per-piece premium below 50 units and print it as its own line on the quote. In total cash, 30 is 6% of the units but nowhere near 6% of the money, because the fixed lines don't shrink. So the number that matters isn't the unit price. It's how many you'll sell at full price this season, and what the leftovers do to your bank account.

Here is where the gap comes from, and where some of it quietly comes back to you.

Why 30 pieces costs more per unit

Two kinds of cost go into an order. One scales with quantity: fabric, trims, minutes on a sewing machine. The other one doesn't. Pattern making, grading, mock-ups, one final sample, and getting a line ready to run cost the same whether the line then makes 30 garments or 500.

Do the division yourself. Call the fixed lines F. At 500 you divide F by 500. At 30 you divide F by 30. If F comes to $1,500, that's $3 a piece at 500 and $50 a piece at 30. Same work, spread sixteen times thinner. That one division explains most of the difference between the two quotes in your inbox.

We won't hand you a universal F, and anyone who does is guessing. Setting up a lined blazer with a chest canvas is not setting up a slip dress.

Cutting and trims don't scale down either

Spreading and cutting a 30-piece lay takes nearly as long as a 100-piece one. The marker still has to be made, the fabric still has to be relaxed and spread, the pieces still get bundled by size. Trim suppliers have minimums of their own, so on a 30-piece run it's normal to buy far more labels, zips or hangtags than you need and shelve the rest.

Fabric has its own minimums

Mills sell by the metre and they have MOQs too. On one of our recent runs, 200 metres of fabric yielded 180 pieces — the rest was width and waste, gone before a single seam was sewn. If a mill's minimum is 200 metres, ordering 30 pieces means you're buying for 180 and storing most of it. Sometimes the smart answer is 30 in one colour plus a second style cut from the same roll.

Quality control doesn't get cheaper at 30

We inspect each piece three times, measure every size, and fit on a live model and a mannequin. That's three inspections per garment whether there are 30 or 500. At 30 pieces our QC cost per unit is roughly sixteen times what it is at 500, and we'd rather show you that line than bury it in the unit price. Plenty of factories solve this by inspecting almost nothing and calling the difference a saving.

Where low MOQ quietly saves you money

Now the other side of the ledger, which in our experience matters more than the unit price.

A low MOQ clothing manufacturer lets you be wrong cheaply. Two things go wrong on most first runs:

• Grading. The sample fits at size 8 and collapses at size 14. Stretch fabric graded with a standard rule set is the worst offender. On 30 pieces you fix the grade and reorder. On 500, you own 500 problems.
• Colour. Two rolls out of the same dye lot don't match, and on silk you can see a spectrophotometer reading that's off by 0.8 Delta E. One bad roll is a phone call. Forty is a write-off.

Then there's the money itself. Our clients retail between $500 and $2,000. Five hundred pieces at $500 is $250,000 of retail sitting in a warehouse before the style has earned anything. At $2,000 it's a million. The premium you pay on 30 pieces is insurance against precisely that.

Three times we talk people out of 30

We do this regularly. It's cheaper for everyone in the long run.

• You're reordering a style that already sold. Below 50 pieces you pay the per-piece premium for nothing. Fifty to 150 per style is where our pricing actually makes sense.
• The fabric minimum forces your hand. If you're buying 200 metres anyway, make the 180 pieces.
• Setup dominates the garment. Heavy draping, sequins, complex bonding — the fixed cost per style is high, so spreading it across 30 is waste. Get to 50 minimum.

And one thing we simply don't do, at any quantity: hand-beading. We say so on the first call rather than quoting it and disappointing you in week six.

What we put in writing

You get a line-by-line quote within 24 hours of a tech pack. No tech pack? Send reference photos and your measurements and we'll build one — digital pattern, full size grading, the lot.

The quote lists fabric, cut-make-trim, trims, sampling and freight as separate lines. Sampling is charged separately and deducted from the bulk order once you place it. The minimum, 30 to 50 per style per colour, goes in writing, with the below-50 premium as its own line so you can see exactly what the flexibility costs.

Timing matters here too, because it decides whether 30 or 500 is even possible. Two to three mock-ups in 14 to 21 days, a final sample in 28 to 42 days, bulk production 6 to 10 weeks. If we're sourcing the fabric, add another 2 to 4 weeks.

Ask yourself how many pieces of one colour you can realistically sell at full price before the season turns. If that answer is 40, a garment manufacturer quoting you 500 isn't doing you a favour.

Nobody sends you a quote with the real cost per garment on it. We write quotes for a living and even ours has lines that surprise people on the first read.

What's missing from most of them? Four things: fabric waste, the second and third sample, freight, and the cost of fixing a mistake after you've signed off. 200 metres of cloth does not make 200 garments — on one recent style it made 180. Add those four to the cheap quote and the ranking usually flips.

Below is how we read a quote sheet, and why two factories can come back three times apart on what looks like the same dress.

The six lines, and what each one hides

On most of our styles fabric is the biggest single line and cut-make-trim is the second. The rest tend to get rounded away, and that's exactly where the trouble lives.

Fabric

The largest line and the least comparable one. Silk has three numbers you have to ask for before anybody can quote it: momme weight, grade, and weave. Silk blend is not a specification. Our wool comes from Biella (Super 120s), silk from Suzhou, cupro from Osaka — name those and a mill can match them, say wool fabric and you'll get whatever was cheap that week.

Fabric pricing also moves with quantity in a way brands don't expect. A mill sells you 60 metres for a 30-piece run at one rate and 1,000 metres at quite another.

Cut, make, trim

The labour. Per piece it barely changes between 30 and 500 units. What changes is how much fixed setup is sitting inside it — pattern, grading, the mock-ups, and the line being made ready.

Trims

Zips, buttons, interlining, labels, hangtags, shoulder pads. Cheap quotes quietly substitute these after approval, which is why we list them by name and check them again at packing.

Sampling

Charged separately, then deducted from the bulk order once you place it. Ask whether the factory does the same, because some keep it.

Quality control

Often not a line at all, and that's the tell. Ours is three inspections per piece, AQL 1.0 to 2.0, measurement at every size, a fit on a live model and a mannequin, and a video record of the batch before it leaves.

Freight

In our quote, named. Sea is cheap and slow, air is fast and brutal. Air freight to cover a late fabric booking can easily cost more than the sampling did — which is why the fabric calendar matters more than the sewing calendar.

The lines brands forget to budget

• Waste and width. 200 metres made 180 pieces. Ten percent gone before cutting even finished.
• Sample rounds two and three. Two to three mock-ups is normal, not failure. Budget 14 to 21 days for them, then 28 to 42 for the final sample in the correct fabric.
• Fabric lead time. Add 2 to 4 weeks if the factory is sourcing. Most late deliveries are fabric delays, not sewing delays.
• Rework. Sign off too early and you pay for the batch twice. It's the single most common way a season goes wrong.
• Colour matching. Same dye lot, two rolls, two colours. On silk, 0.8 Delta E is visible to the eye.
• Grading. Stretch fabric put through a standard grade rule collapses at the upper sizes. Fix it on paper, not in production.

That list is the real garment production cost. None of it appears on a quote, all of it lands on you.

If you want a fair comparison, send the same written spec to three factories and ask each one for the same six lines. Any custom clothing manufacturer should manage that, and two of the three will. The third sends a single number with a friendly note, and that reply is itself the information you needed — they either can't split it or don't want you looking. We would rather lose the job than quote one round figure, because a round figure always gets re-quoted the moment something changes.

Why two quotes come back three times apart

We lose jobs on price and then get the call six weeks later. The gap is almost never one thing. It's usually five.

Fabric first — same name, different grade. Then QC: a factory doing one inspection on a handful of pieces is not doing what a factory doing three per piece is doing, and shouldn't cost the same. Then sampling — one round in whatever cloth is on the shelf, versus two to three mock-ups plus a final sample in the actual fabric. Then MOQ, because a factory built around 300 to 500 units per style prices for that volume, so its number for 30 pieces is either a favour or a trap. And finally what's excluded: the lines that aren't on the page because nobody asked.

One more, and it's the nastiest. The fabric supplier ships you a confirmation sample, you approve it, and the bulk goods that arrive are not the same cloth. That has happened to us. The only defence is keeping a sealed approved sample and taking a spectrophotometer reading when the goods come in, which is why we do both.

So when a Guangzhou garment manufacturer comes in at a third of everyone else, the useful question isn't why so cheap. It's which of those five they took out.

How we quote, and what we won't quote

Tech pack in, itemised quote back within 24 hours, every line separate. No tech pack? Send reference photos and your measurements and we'll build one — digital pattern and full size grading included. At the end of the job you get a written report on what went right and what didn't, which is the part most clothing manufacturer relationships skip.

Two things we won't do. We won't quote a silk dress until momme, grade and weave are settled — we've been burned by that sentence more than once. And we don't take hand-beading, at any quantity.

A cheap quote isn't a lie. It's a different garment. Ask which one you're buying.

Made to order sounds like the cautious, slower option. Run the cash numbers and it's often the reason a three-person label is still trading in year three.

Stock means paying for everything up front — fabric, 6 to 10 weeks of production, storage — and then being paid back slowly as it sells. Pre-order means being paid before or while you make it. If the style is new, or the market is new, 30 pieces beats 500 even at a worse unit price. That's the argument, and the rest of this is the arithmetic behind it.

What inventory actually eats

Three things, and only the first one is on the invoice.

Cash goes out early. Fabric takes 2 to 4 weeks if the factory is sourcing it, the final sample takes 28 to 42 days, and bulk production takes 6 to 10 weeks. Call it up to 14 weeks from the day you commit to the day goods are in hand. Every dollar is spent months before the first sale.

Cash comes back slowly, and never completely. Our clients retail between $500 and $2,000. Five hundred pieces at $500 is $250,000 of retail that has to clear at full price before the style earns a cent. At $2,000 it's a million. Whatever doesn't clear gets marked down, and that markdown comes straight out of next season's budget.

Then the cost nobody calculates: while that money sits on a shelf, it isn't buying fabric for the next drop. The brands we work with are 1 to 50 people. One wrong 500-piece buy can be the whole year.

The 30-piece version of the same season

A small batch clothing manufacturer charges more per unit — the garment production cost per piece really is worse at 30. That part you already know. What you're buying with the premium is the right to be wrong cheaply.

Make 30 in the hero colour and 30 in the second. Sell them. If size 14 is the one that keeps coming back with a fit complaint, you now know before you've committed 500 units to it. If two rolls from the same dye lot don't match, you've found it on one roll instead of forty.

There's a quality argument too. We inspect each piece three times, at AQL 1.0 to 2.0, measuring every size, with a video record of the batch. That's a fixed amount of work per garment. At 30 pieces it's affordable and it's thorough. At 500, a factory under pressure to hold the unit price is under real pressure to inspect less — and that's where a lot of volume savings quietly come from.

Our first-sample pass rate is 98%. That isn't a quality boast. It means sampling usually doesn't swallow an extra six weeks, which is the thing that makes a pre-order calendar workable at all.

When 500 is the right call

• You've already sold this style once and you're refilling a size curve you trust.
• You tested a colour at 30 and it sold out in three weeks. Reorder at 100 to 150 — that's the band where our pricing makes the most sense.
• The fabric forces it. 200 metres yielded 180 pieces on one of our runs. If the mill's minimum is 200 metres, you're buying that much whether you make 30 or 180.
• You have a fixed delivery date and no runway to run pre-order properly. Pre-order only works once sampling is done: two to three mock-ups in 14 to 21 days, then the final sample in 28 to 42 days.

What pre-order demands from you

It isn't free money, though. Pre-order only works if you've done the boring part early.

Sampling has to be finished before you open the window: two to three mock-ups in 14 to 21 days, then the final sample in the correct fabric in 28 to 42 days. Only then can you photograph the real garment and take orders against it. Rush that and you're selling a picture of something that doesn't exist yet, which is how small labels end up paying for air freight they never budgeted.

You also have to be straight with customers about the wait. Bulk production is 6 to 10 weeks, plus 2 to 4 weeks if we're sourcing the fabric. Put those numbers on the product page. The shoppers who mind were never going to buy a small-batch piece anyway, and the ones who stay will wait, because they're buying something made for them rather than pulled off a shelf.

And you have to commit to a size split before we cut. That's the one number you can't correct afterwards, so use whatever sales data you already have, even if it came from a weekend market stall.

How we'd split a first buy

If you're a studio of 1 to 10 people launching into a new market, this is the plan we keep repeating. Thirty to 50 pieces in the hero colour, 30 in the second, and hold the third colour for a reorder. Put the reorder date in the calendar before you launch, not after you've sold out.

Below 50 pieces there's a per-piece premium, and we show it as its own line instead of averaging it away. You should be able to see what the flexibility costs and decide for yourself whether it's worth it.

Watch one more number while you're at it: how long your cash is tied up. Fabric 2 to 4 weeks, sampling 28 to 42 days, production 6 to 10 weeks. If the money has to sit there for 14 weeks before the first sale, the unit price is not your problem.

Ask yourself what $250,000 of unsold coats buys you. Then ask what 30 would have told you for a fraction of that.

Everyone asks how to find a clothing manufacturer they can trust, then skips the part that actually protects them: the factory audit.

Verify three things before you send a deposit. The workshop address matches the business licence. The QC method is written down in numbers — an AQL level, inspections per piece, measurement at every size. The MOQ appears on the quote. If any of those three is vague, nothing else in the conversation matters yet.

We're a Guangzhou garment manufacturer and we get audited most weeks. This is the checklist we'd want used on us, including the parts buyers usually forget to ask.

Seven red flags

1. The bank account name doesn't match the licence

Cross-check the name on the business licence against the name on the bank details. A trading company will happily take your money under a different entity, and when something goes wrong you're chasing a person, not a registered workshop. Pay the company that makes the goods.

2. You never see the cutting floor

Ask for a live video call from the cutting tables, not a showroom walkthrough. Showrooms are rented. Cutting tables, fabric store and a finishing section are not, and no factory that owns them refuses to show them. If you get a conference room and a catalogue, you're hiring a middleman who will subcontract to whoever is cheapest that month.

3. MOQ is no problem but never in writing

Every factory can say yes to 30 pieces. A real small MOQ factory quotes a number per style, per colour, with the fixed costs and the below-50 premium broken out as separate lines, and it goes in the contract. If the MOQ isn't written down, it doesn't exist.

4. They can't tell you where the fabric comes from

Ask which mill. A factory that genuinely handles good cloth answers in names and cities — Biella for wool, Suzhou for silk, Osaka for cupro. We have many fabrics means they'll buy whatever is available after you sign, and that's how an approved sample turns into different bulk goods.

5. The delivery date is promised before the mill confirms

Fabric sourcing takes 2 to 4 weeks and production takes 6 to 10 weeks. A four-week promise on a style that needs sourced fabric is fiction, and it's the most common lie in this industry. Ask what the mill has confirmed in writing before you accept any date.

6. QC is described as very good

Numbers, not adjectives. Which AQL level? How many inspections per piece? Is every size measured, or only the sample size? Is there a fit on a live model and a mannequin? For reference, we run three inspections per garment at AQL 1.0 to 2.0 and keep a video record of every batch. A factory that can't answer in numbers is describing a hope.

7. One sample round, wrong fabric, pressure to sign now

The rush is the tell. Normal development is two to three mock-ups in 14 to 21 days, then one final sample in 28 to 42 days using the actual fabric. A factory pushing you to approve a mock-up made in a substitute cloth is setting up a rework bill, and you'll be the one paying for the whole batch twice.

What to look at when you visit

Turn up on a working day, not a scheduled open house, and walk these in order:

• Cutting tables. How many, and whose fabric is on them. Other people's work tells you the real level of their clients.
• The sample room. Patterns filed and dated, or in a pile on the floor?
• Sewing lines. Count the operators, and check they're making your category. A line that lives on t-shirts will struggle with a structured blazer.
• Finishing and pressing. This is where cheap work shows — untrimmed threads, rushed pressing, no final inspection table.
• The fabric store. Are rolls labelled with lot numbers, or unlabelled?
• The QC table. Measuring tapes, size charts, a light box, and someone actually measuring each size.

Then ask two questions that aren't on any standard audit form: show me the video record from the last shipment, and show me a garment you rejected last month. A factory with zero rejects is lying to you. Bring your own size chart and pick a finished piece off the line at random, then measure it yourself.

Five questions for a small MOQ factory

• What is your MOQ per style per colour, in writing?
• How many mock-up rounds, in what fabric, and how long? (Ours: two to three, 14 to 21 days, then a final sample in the correct fabric in 28 to 42 days.)
• How do you inspect, at what AQL, and do you measure every size?
• What's your first-sample pass rate? (Ours is 98%, and we'll tell you what the other 2% looked like.)
• What don't you do? We don't take hand-beading at all, and we'd rather say that on day one.

How to tell whether an AQL report is real

We have seen reports copied from the previous year, with the date changed and nothing else. Four checks catch most of it.

First, the batch number and date on the report have to match your purchase order and the actual ship date, not a generic template. Second, ask which inspection level and which table row they used — if nobody can answer, the number was typed rather than counted. Third, look for a measurement table: spec, tolerance, and an actual reading at every size, not one row for the sample size. Fourth, ask for the video record of that batch and check the date on the file.

One more, if there's colour in the order: ask for the spectrophotometer reading. On silk, 0.8 Delta E is visible, and a report with no colour data at all hasn't been near the goods.

And treat 0% defects with suspicion unless every line above it is filled in. Perfection on a page with no measurements attached is marketing.

Any factory can pass an audit it sees coming. Show up on a Tuesday with your size chart in your pocket and ask to see a rejected garment — the answer tells you more than the report does.

AQL sounds like a technical detail. It isn't. It's the single number that decides how many defective garments can legally end up in your customer's hands.

AQL stands for Acceptable Quality Level. AQL 1.0 means: on a random sample, we accept at most 1 defect per 100 units. AQL 2.5 allows 2.5. Doesn't sound like much of a gap. On a 500-piece order it's the difference between 5 faulty pieces and 12 — and if you retail at $500, that's $3,500 of returns sitting in the difference.

Why most factories default to 2.5

Because it's cheaper to run. Fewer rejects means fewer remakes, and remakes come out of the factory's margin, not yours. Nobody says this out loud in a quotation email. They just write "AQL 2.5 standard" in a font small enough that you skim past it.

We run AQL 1.0–2.0. And we inspect every single garment, not a sample. That second part matters more than the number, honestly. A perfect AQL on a 10% sample tells you nothing about the other 90%.

What the three stages actually check

Here's the part most brands never see, because it happens behind a door.

Stage one, incoming fabric. Before anything is cut. We measure width, check for holes and shading, and — this is the one people skip — test shrinkage. A fabric that shrinks 4% in the wash will produce garments that fit fine on day one and don't fit after the customer's first laundry cycle.

Stage two, inline inspection. Mid-production, on the line. This is where construction errors get caught: a seam that's twisting, a pleat that's sitting 3mm off, a facing that's pulling. Catch it here and you rework 20 garments. Catch it at the end and you rework 400.

Stage three, finishing. Measurements, size by size. Not "the medium looks right" — actually measuring a size 8 and a size 14 against your spec, because those two almost never behave the same way. Then live model try-on and a mannequin check. Then the whole shipment gets filmed before it goes in the box.

How to tell if an AQL report is real

We've seen reports copied from a previous season with the date changed. It happens more than you'd think. Three things to look for:

• Sample size must match the AQL table for your order quantity. A 500-piece order at AQL 1.0 with General Inspection Level II means 50 pieces pulled, not 20. If the numbers don't match the table, the inspection didn't happen the way they claim.
• Defects should be categorized — critical, major, minor — with actual counts. A report that says "passed" with no counts is a stamp, not an inspection.
• Ask for the measurement sheet, not just the pass/fail. Real inspectors record numbers. Fake ones record conclusions.

If a factory gets defensive when you ask for the raw sheet, that's your answer.

The honest caveat

AQL 1.0 costs more and takes longer. Three days of inspection time on a 50-piece order is a bigger relative hit than on a 1,000-piece order — which is one reason small MOQ clothing manufacturers quietly push you toward looser standards. Ask what you're actually getting. We'd rather quote you the real number than win the order and argue about defects later.

Not every brand needs AQL 1.0. If you're making basic jersey at $60 retail, 2.5 is fine and we'll say so. At $500 plus, with beading or silk, it isn't.

The final payment is the only leverage you have left. Once it's wired, a container is on a boat and your complaints become a negotiation instead of a checklist.

So what should actually be checked before that payment goes out? Not "quality" in the abstract. Specific, measurable things.

Seams: three failures that matter

Stitch density. Count stitches per inch. Too few and the seam pulls open under stress. Too many and you perforate the fabric — this is a real problem on silk and fine synthetics, where a dense needle line creates a tear line. Most dress manufacturers land at 8–12 SPI for woven. If a factory can't tell you their SPI, they're not controlling it.

Seam slippage. Pull the seam sideways. On loosely woven fabrics the yarns slide apart and you get a gap that no amount of re-stitching fixes. This is a fabric problem, not a sewing problem, and it should have been caught at incoming inspection — but check it again on finished garments, because stress points differ.

Puckering. A seam that looks fine flat but ripples when you run a hand along it. Usually feed tension, occasionally the wrong needle for the fabric weight. Looks cheap, feels cheap, and customers notice before they can name it.

Tests worth paying for

You don't need every test on every order. But on a first run with a new factory, or any order over a few hundred pieces, three are worth it:

• Wash test. Shrinkage and color fastness. Your customer will wash this garment. Find out now what happens.
• Rub test (crocking). Rub a white cloth on the fabric. If color transfers, it will transfer onto the customer's bag, their coat, the chair at the restaurant. Particularly relevant on saturated dyes and dark denim.
• Pull test. On buttons, snaps, and any attached trim. A button that comes off in week two generates the angriest kind of review — the kind that says the brand doesn't care.

Documents to have in hand before the final payment

Ask for these and read them. Don't file them.

• Measurement sheet, size by size, with tolerance noted
• Defect count by category (critical / major / minor)
• Packing list that matches your carton markings
• Fabric test results, if you ordered them
• Photos or video of the packed shipment

We video every shipment before it closes. Not as a marketing thing — because the alternative is arguing about what happened six weeks later, over email, in a different time zone, with no evidence.

One thing nobody warns you about

Check the cartons, not just the garments. We've taken over projects where the clothes were fine and the packing was a disaster — wrong size ratios per carton, mixed colorways, no polybags on silk. Your boutique receives ten boxes and has to sort all of it before they can put anything on a rail. That's not a quality defect. It's still a reason they don't reorder.

Two rolls. Same mill, same dye lot number, delivered on the same day. Hold them side by side and they're different colours.

This is not a scam and it's not incompetence, most of the time. It's physics, plus a supply chain that doesn't care about your launch date. But it is fixable, and the fix starts with a number.

Delta E, without the maths

Delta E measures how far apart two colours sit in a standard colour space. It's a distance, not a percentage. Rough working guide:

• Under 1.0 — you can't see it. Nobody can, under normal lighting.
• 1.0 to 2.0 — visible to a trained eye side by side. Acceptable for most production.
• Above 2.0 — visible to anyone. On a garment with two panels cut from different rolls, this is the complaint you get.

We hold 1.0 as the working limit against the approved lab dip. Not because 1.0 is magic, but because by the time fabric reaches the cutting table, colour has already moved once.

Why readings drift

Spectrophotometers are measurement instruments, and measurement instruments drift. Lamp output changes over time. The aperture gets dusty. Someone drops it. Then one morning it reads 0.8 off from where it read in March, and every approval since March is suspect.

We caught this in our own setup — a 0.8 shift over roughly three months, which is enough to pass fabric that should have been rejected. Nothing was "wrong" with any single reading. They were all consistently wrong in the same direction.

The fix is boring: a certified reference tile, checked at the start of each session, with a documented calibration log. Quarterly professional recalibration. If your factory can't produce that log, their Delta E numbers are decoration.

Where colour actually goes wrong

Three places, in order of how often they bite:

Lab dip to bulk. The lab dip is mixed in a beaker under ideal conditions. Bulk is 400 metres in a vat with different water, different timing, a different operator. They will not be identical. Bulk should be approved against a Production Standard, not the original lab dip.

Roll to roll within bulk. Dyeing is a continuous process with a beginning, middle, and end. The tail of the run differs from the head. Mills know this. They may not tell you which roll came from where.

Lighting. Metamerism — two fabrics matching under store LED and diverging in daylight. Check approvals under at least two light sources. This one generates the worst arguments, because both parties are telling the truth.

What to actually do about it

Approve a physical sealed sample, signed, dated, kept by both sides. Test incoming rolls against it before cutting. Number the rolls and record which garment came from which roll, so if a complaint arrives six months later you can trace it. And reject early — a shade problem found before cutting costs you a re-dye. Found after sewing, it costs you the whole run.

On one silk order this discipline saved about $58,000. Not through cleverness. Just by checking before cutting instead of after.

Most production disasters are decided before a single garment is cut. They're decided at the fabric stage, by someone who didn't know there were six separate places to lose money.

1. Mill minimums don't care about your MOQ

You found a factory that will make 30 pieces. Congratulations. The mill still wants 1,000 metres.

This is the single most common surprise for new brands. Garment MOQ and fabric MOQ are two different numbers set by two different businesses. Fabric mills run dye vats; a small vat is uneconomical, so minimums sit high regardless of how few garments you're making.

Three ways out: buy stock-supported colourways (mills keep greige and neutrals), combine with another brand's run, or accept a surcharge for a small dye lot. The third is usually cheapest at 30–50 pieces, and it's why a low MOQ clothing manufacturer quotes a higher per-metre price than you expected.

2. Lead time starts when the mill confirms, not when you order

Fabric takes 2–4 weeks on top of production. If your factory promises a delivery date before the mill has confirmed your order, they're guessing — and you'll absorb the miss.

Sequence it properly: approve lab dip, confirm bulk, receive and inspect fabric, then book production. A factory that books the line before fabric lands is betting with your launch date.

3. Substitution is the quiet one

You approve a 19-momme silk charmeuse. What arrives is 16-momme and feels thinner in the hand. Nobody swapped it maliciously. They ran out, and something similar was available, and the deadline was close.

Catch it at intake, not at cutting. We weigh and measure every incoming roll against the approved spec: weight per square metre, width, composition. On one order a wool "Super 120s" came in as a blend — real wool, wrong grade, and invisible until tested.

4. Width eats your piece count

You ordered 200 metres. You get 180 pieces, not 220.

W usable width, not nominal width, determines consumption. A fabric sold as 140cm that measures 114cm between selvedge flaws changes your marker entirely. Add shrinkage — a 4% shrink means your cut pieces are smaller than the pattern assumed — and the gap widens again.

Ask for usable width in writing. Then add 3–5% for cutting waste on top of that.

5. Origin certificates are easy to fake

"Biella wool" is a marketing phrase as often as a fact. Mills can and do supply documentation for cloth they didn't weave.

If origin matters to your brand story — and for luxury positioning it usually does — verify at intake and ask for mill documentation that traces the lot. We source wool from Biella, silk from Suzhou, and cupro from Osaka through relationships built over years, precisely because traceability can't be established after the fact.

6. Deadstock and "sustainable" need the same scrutiny

Deadstock is genuinely better for the planet and genuinely harder to work with. You get one shot at the quantity available. There is no re-order, no second dye lot, and often no composition documentation.

And "sustainable" on a fabric label means nothing without a certificate — GOTS, Oeko-Tex, GRS. Ask for the certificate number and check it with the issuing body. We've seen stock sold as certified with no valid certificate behind it.

The part where we're honest

We'd rather you supply your own fabric than fight about ours. Many clients nominate their own mills, and we handle receipt, inspection, dye-lot matching, and inventory. If your supplier fails, we switch to our Guangzhou network — but that fallback has a cost, in time if not in money.

Either way: inspect before cutting. Every time.

Sample fits beautifully at size 8. Bulk arrives and size 14 is unwearable. Same pattern, same factory, same fabric.

This is the most common failure in small batch production, and it's almost never a sewing problem. It's grading.

Grading is not scaling

Software can scale a pattern. Type in a percentage, get a bigger version. Fast, free, and wrong for anything past one or two sizes either side of the sample.

Bodies don't scale proportionally. A size 14 isn't a size 8 multiplied by 1.2 — the proportions change. Bust to waist ratio shifts. Shoulder slope changes. Rise behaves differently from inseam. A grade table that applies one rule across all points will drift further the further you go from the base size.

Hence: fine at 8, acceptable at 10 and 12, broken at 14.

The drift is measurable

We check grade increments point by point, not as a single number. A grade increment table should specify, per measurement point, how much is added per size — and those increments should not be uniform.

Where it usually breaks:

• Armhole and sleeve crown. Uniform grading shrinks the armhole relative to the body. Size 14 ends up tight under the arm even though the chest measures right.
• Rise on trousers. Grade rise and inseam with the same increment and the seat pulls on larger sizes.
• Ease. This one is subtle. Ease — the gap between body measurement and garment measurement — is a design decision, not a constant. Grade it as a constant and a size 14 wears like a size 38 with the same nominal spec.
• Extremes on stretch. A standard grade table on elastane fabric compounds error. The fabric recovers differently at different widths, so the same increment produces different results across the size run.

Size sets exist for this reason

One mock-up tells you nothing about grading. You need a size set — samples at the base size and at both extremes, made in the actual production fabric, after the grade table is applied.

It costs more and it takes longer. It also catches the problem when it's still a pattern adjustment, not 400 finished garments.

What to measure

Agree a measurement standard before sampling, in writing. Which points, how measured, tolerance per point. Sounds pedantic until you discover your "across chest" is measured flat and the factory's is measured on the round, and every garment is 2cm "wrong" in both directions depending on who's holding the tape.

We use a fixed point list and record every size on every order. It's a boring document. It ends arguments.

Inclusive sizing is a separate discipline

If you're running past a straight size range into extended or inclusive sizing, grade from a separate block. Extending a size 8 block to a size 22 rarely works, because the body proportions diverge too far from the base. It's more pattern work up front. It's the difference between a size range that sells and one that generates returns.

We'll tell you if a range you've planned is going to fight you. Better in week one than week ten.

Three rounds in and the fit still isn't right. Everyone's frustrated, the calendar's shot, and nobody can say exactly what went wrong.

We've been on both sides of this. Here's where it usually breaks.

1. The fit block was never settled

This is the big one. A fit block is your base pattern — the sloper everything else is built from. If it isn't locked, every style inherits its problems, and each round is compensating for a foundation issue instead of refining a design.

Signs your block isn't settled: the same fit complaint appears across unrelated styles. Shoulders too narrow on a dress, a blazer, and a coat. That's not three problems. That's one block problem showing up three times.

Fix the block first. It feels like it's slowing you down. It's the opposite.

2. Fit comments arrive as screenshots

"Slightly tight here" with an arrow drawn on a phone photo. Then a separate email about the hem. Then a call about the sleeve.

By round three nobody has the full list in one place, and two of the comments contradict each other. One round says take in the waist; the next says the waist is fine but the hip needs room. Both get applied. Now it's worse.

Put every comment on one numbered sheet, per round, with measurements attached. "Waist -1cm" is actionable. "Feels a bit tight" is not.

3. Sign-off happens on the wrong sample

The single most expensive mistake in sampling: approving on a single-size mock-up in a substitute fabric.

Substitute fabric behaves differently. It drapes differently, it shrinks differently, and it hides or exaggerates fit issues that the real fabric will reveal. Approve a garment in charmeuse, cut it in crepe, and you've approved something that doesn't exist.

Then there's single-size sign-off. You approved a size 8. Bulk runs 8 through 14. Nobody checked the extremes. See the grading article for how that ends.

What a good round sequence looks like

Two to three mock-ups in similar fabric, refining fit and construction. Progress photos at each stage so you're not guessing. Then one final sample in the correct fabric, with a size set if the run spans more than three sizes.

Mock-ups take 14–21 days. The final sample 28–42 days. Our first-sample approval rate is 98% — not because we're geniuses, but because the process forces the decisions to happen in the right order.

Sealed samples and first-piece checks

Before bulk starts, seal the approved sample. Both sides sign it, both sides keep one. This is your reference if there's a dispute.

Then, when the line starts, pull the first piece off and check it against the sealed sample before running the rest. Most factories skip this. It's the cheapest quality gate in existence — one garment checked instead of four hundred made wrong.

The bit nobody likes hearing

Sometimes the design can't be produced the way it's drawn, at the volume you want, at the price you need. Better to hear that in round two than round five. We say it early, because the alternative is a sample that sort of works and a bulk order that doesn't.

Send a sketch and a mood board, ask for a quote. You'll get one. It'll be wrong.

Not dishonestly wrong — just built on assumptions, because the things that determine cost weren't specified. Then the sample comes back, it isn't what you pictured, and you've spent three weeks discovering that your construction and the factory's construction were two different garments.

What's actually missing from most tech packs

Not the sketch. The sketch is usually fine. What's missing is the stuff underneath:

• Measurement spec with tolerance. Every point, every size, and how much variation is acceptable. "Chest 92cm" is not a spec. "Chest 92cm, ±1cm" is.
• Construction notes. Seam type, seam allowance, how the facing is attached, whether the hem is blind-stitched or topstitched. Two factories can read the same sketch and build it two ways, both reasonable.
• Fabric and trim, fully specified. Composition, weight, width, and — for trims — the actual supplier code. "Gold button" generates a question. "Button, 15mm, style #4412 from supplier X" doesn't.
• Grading instruction. Base size, size range, and whether you have a grade table. This decides whether your size 14 fits.
• Labelling and packing. Where the label goes, what it says, how it's attached. Sounds trivial. It's a line item and a common last-week problem.

What it costs you

Every unspecified item becomes a decision someone else makes. Usually a reasonable one. Sometimes not yours.

Then it surfaces at sampling, where fixing it costs a round instead of an email. A round is 14–21 days. Three rounds is most of a season for a small brand.

If you don't have one

Fine. Send reference images, sketches, and measurements and we'll build the tech pack. It's part of the process, not an upsell.

What you get back is a document you own and can take anywhere — which means the next quote you request, from anyone, will be faster and more accurate. That's worth having even if you never work with us.

A note on over-specifying

The opposite failure exists too. Twenty pages on a 30-piece order, with tolerances tighter than the fabric can physically hold. That slows everyone down and adds cost for precision nobody will see.

Specify hard on the things that affect fit, cost, and consistency. Leave the rest to the people who do this daily.

Ask ten factories what makes them "high-end" and ten will say quality. It's not an answer.

Here's what actually differs, in ways you can verify.

Rejecting the first sample

A factory that approves everything is either not looking or not capable. We turn down work — including hand-beading projects, which we don't take at all — and we push back on first samples when the construction won't hold at volume.

That sounds like a sales disadvantage. It's the opposite. A sample that can't be reproduced consistently will fail in bulk, and you'll blame the factory and the factory will blame the spec.

Composite techniques in one place

Draping, embroidery, pleating, sequin work — most factories do one and subcontract the rest. Each handoff is a chance for the design to drift, and each subcontractor has their own minimum and their own lead time.

We handle composite construction in-house. Fewer handoffs, fewer places for a beading pattern to shift by 3mm between the sample and the run.

Sequin and bead alignment is a jig problem

Not a skill problem, mostly. Aligned sequins across a size run need a physical jig and a documented placement, checked per size. Without it, the pattern that looked deliberate on the sample becomes approximate by size 12.

Anyone can hand-place beads beautifully on one garment. Ask how they hold it across 300.

Three inspections, every garment

Not a sample. Every piece: incoming fabric, inline on the line, finishing. Size-by-size measurement, live model try-on, mannequin check, AQL 1.0–2.0.

Spot-checking is faster and cheaper and it's what most factories do. It also means a defect rate you'll discover through customer returns rather than through your own QC.

Documentation you can actually use

Every shipment gets filmed before it closes. Measurements are recorded per size. After the project you get a written debrief — what worked, what to change next season.

That last one sounds like a nice-to-have. It's the thing that makes season two cheaper than season one.

When you don't need any of this

Honestly? If you're making basic jersey at $60 retail, we're the wrong factory and the premium makes no sense. AQL 2.5, one sample round, a standard factory — that's a rational choice for that product.

Our clients retail at $250–$1,000, where a customer will run their hand along a seam and decide. That's the gap we're built for.

"Silk" is not a specification. Two mills will send you two completely different fabrics for the same request, both legitimately called silk, and one of them will fall apart in the first wash.

Three numbers before you accept a quote.

Momme

Momme measures silk weight — roughly, the weight in pounds of a piece 45 inches wide by 100 yards. Higher means heavier and generally more durable.

Working ranges: 12–16 momme for linings and light scarves. 16–19 for blouses and flowing dresses. 19–22 for structured pieces that need body. Above 22 for tailoring.

Here's the trap. A mill quotes you 19 momme. The fabric arrives and it's beautiful, but it's lighter in the hand than you expected. It's not mislabelled — momme tolerances are real, and a fabric at the bottom of the 19 range behaves differently from one at the top. Specify a range in writing, then weigh the delivery.

Grade

A, B, C. Grade reflects how much of the filament can be used continuously — long filaments produce smoother, stronger cloth; short ones produce slubs and weak points.

Grade A is what you want for anything where the surface matters: charmeuse, anything with sheen, anything photographed close. Grade B is fine for linings or where slubs read as character. Grade C is where the fabric pills and the seams slip.

You cannot reliably grade silk by looking at it in a photo. Handle it, or get a sample.

Weave

Charmeuse, crepe de chine, habotai, georgette, dupioni, organza. Same fibre, completely different behaviour — drape, sheen, shrinkage, how it takes a needle.

Crepe de chine shrinks differently from charmeuse. Georgette needs a different needle and a different feed tension. A factory that's only run cotton shirting will struggle with charmeuse, because charmeuse moves under the presser foot in a way woven cotton doesn't.

Then the dye lot

Silk takes dye unevenly compared to cotton, so lot-to-lot variation is more pronounced. Two rolls from the same dye lot can differ. Test each incoming roll against a sealed standard before cutting.

And buy enough the first time. Silk re-orders are the worst kind of gamble — the mill's next lot will be close, and "close" on a garment where panels sit side by side is visible.

Small quantities, worst lots

One more thing. If you're buying 60 metres against a mill minimum of 1,000, you are at the back of the queue for quality control. Short runs get scheduled at the end of a dye cycle, when the bath is least consistent.

It's not malicious. It's how the economics work. Which is why we inspect every metre at intake, and why we'll tell you upfront that small-quantity silk carries more colour risk than the same order in wool.

Production takes 6–10 weeks. Somehow your order took 16. Neither number was a lie — they were measuring different things.

What the 6–10 weeks actually covers

Cutting, sewing, finishing, inspection. That's it. Fabric isn't in there. Sampling isn't in there. Shipping definitely isn't.

Add them up and the real timeline looks like this:

• Tech pack and quoting — a few days to a week, longer if the spec is thin
• Mock-up rounds — 14–21 days each, and most brands need 2–3
• Final sample — 28–42 days in the correct fabric
• Fabric sourcing — 2–4 weeks, starting when the mill confirms
• Production — 6–10 weeks
• QC and packing — 3–5 days
• Shipping — 5–8 days by air, 18–25 by sea

Realistically five to six months from first email to landed goods, if you're sampling properly.

Where it actually slips

In our experience, in this order:

Fabric confirmation. The biggest single variable. A factory that books the production line before the mill confirms your fabric has committed to a date it doesn't control. When the mill slips, everything slips. We confirm the date only after the mill does — which means we occasionally quote a later date than a competitor. That date holds.

Sample approval delays on your side. Not a criticism, just arithmetic. A week to review a sample is a week off the back end. Two rounds of that and you've lost a fortnight.

Scope changes mid-production. Changing a trim after cutting starts is not a small change. It's a re-order and a re-schedule.

Re-dyes. If a lab dip fails, that's another 1–2 weeks. It happens. Plan for one.

How to lock it

Get the date in writing, with the dependencies named. Not "delivery by October" but "production starts when fabric is confirmed and inspected; 8 weeks from that date."

Ask what happens if fabric is late. A factory with a real answer has done this before.

Then track it. We give live order tracking, not because it's a feature but because "where's my order" emails are a symptom of nobody knowing.

Sea or air

Air is 5–8 days and costs several times more. Sea is 18–25 days. On 30 pieces, air is usually worth it. On 500, the maths flips.

Factor duties and customs into both — the shipping quote isn't the landed cost, and this is where a lot of first-time importers get surprised.

Finding an evening wear manufacturer for the Gulf is easier than finding one that understands the market. Abayas, kaftans and gowns sold in Kuwait, Saudi Arabia and the UAE are judged by rules that don't exist in London or New York.

We make 30 to 150 pieces per style from our workshop in Guangzhou, and a growing share of those orders ship to the Gulf. This is the checklist we use with clients from Kuwait City, Riyadh and Dubai before they commit a deposit — the things that decide whether the season works or the rack sits.

Before picking a clothing manufacturer Middle East: what to verify

Start with the wrong question and every answer after it is wasted. Most labels ask about price first. The useful first question is different.

Ask what the factory has actually shipped to the Gulf, and what it learned there. Shipment history is worth more than any certificate. A factory that has dressed a Dubai bridal season knows the fabric weight that survives Emirates heat, the embellishment density that sells, and the freight lane that clears customs.

Then check the small MOQ. Gulf labels rarely need 500 pieces. They need 30 to 60 pieces that are right. A factory built around 500-piece runs will quote 30 badly or refuse it.

Abaya and kaftan construction: three details that decide the garment

Kaftan and abaya construction looks simple and is not. The shape is generous, which means the fabric does the work.

Check the fall. A 12-ounce silk twill kaftan hangs differently from a 9-ounce one, and the difference shows at the hemline. We spec fabric weight in ounces per yard and confirm it on a sealed sample before bulk cutting.

Check the finish on the edge. Hand-rolled hems at 3mm are the standard we hold for silk evening pieces; a machine-rolled edge looks different at arm's length.

Check how the garment is lined. Gulf seasons run long and warm, and a lined bodice changes how a kaftan sits on the shoulder.

Embellishment that survives a Gulf season

Gulf evening wear leans on embroidery, beading and stonework. This is where small MOQ factories in Guangzhou either excel or quietly fail.

Two things matter: density and anchor. Beading density should be specified per square inch — we work at 120 to 160 beads per square inch on bodices — and every bead should be double-anchored so a single thread break doesn't shed a line of stones.

We pull-test anchors at 8N on 10% of a run. That sounds technical; it just means we try to pull a bead off before you wear it, not after.

Embroidery thread should be colour-fast to sweat and heat. A white thread that yellows by the third wear is a Gulf-specific failure we've seen too many times.

Modest fit is a specification, not a style

For Gulf clients, coverage is part of the tech pack, not an afterthought. Write it down or it will be guessed.

Neckline depth, sleeve length, hemline and layering ease all need numbers. We grade for these like any other measurement, and we fit on a live model plus a mannequin at each size.

The honest part: we have delivered a first mock-up with the neckline a centimetre deeper than the client intended. We caught it in review, corrected the pattern, and re-cut. Anyone who tells you this never happens is not looking hard enough.

Colour: the Gulf standard is not the European one

Gulf orders lean on deep jewel tones, golds and blacks — colours where dye lots drift most visibly. This is why we hold every incoming roll to Delta E under 1.0 against an approved standard, checked under three light sources: daylight, shop light and warm light.

Two rolls from the same dye lot can still differ. We test every roll before cutting and reject anything that fails. One bad roll is a phone call; forty is a write-off.

Metallics are their own problem. Gold thread and gold beads shade differently across suppliers, so we seal the actual thread and beads, not a description of them.

Timing: Eid, wedding season and the calendar that matters

Gulf retail has two hard windows: Ramadan and Eid, and the bridal season. Both move every year, and both punish late delivery harder than any market we know.

Work backwards from the window, not from the quote. Sampling runs two to three mock-up rounds in 14 to 21 days, then a final sample in 28 to 42 days. Bulk production is 6 to 10 weeks after fabric confirmation, plus 2 to 4 weeks if we are sourcing the fabric.

For a 30 to 60 piece order, air freight at 5 to 8 days is usually worth the cost. We handle export documents and packing lists for Gulf customs.

Small MOQ factory UAE and the region: what 30 pieces buys you

Gulf labels search for a small MOQ factory in the UAE or a Guangzhou manufacturer that ships there. The region itself has limited small-batch production; most serious evening wear runs still come from China.

Thirty pieces per style per colour buys you three things. First, a real test of the market before you commit to volume. Second, cash that stays in your account instead of in a warehouse. Third, the right to be wrong cheaply — a fit or colour fault costs one roll to fix, not forty.

We add a per-piece premium below 50 units and print it as its own line, so you can see exactly what the flexibility costs.

How we work with Gulf labels

Send reference photos, your measurements, and your target retail price. We build the tech pack if you don't have one, quote line by line within 24 hours, and keep your patterns on file so reorders don't restart.

For series planning, we are also training our first design tool — an AI agent we call A1, coming online soon. You describe your brand positioning, the styles you need, and a colour theme, and it drafts design images to start the conversation. It is still in training, so treat early output as a sketch, not a final line.

If you are sourcing evening wear for Kuwait, Saudi Arabia or the UAE and want a factory that has been through the Gulf season more than once, start the conversation and see how we answer your first three questions.

Cold emails to boutiques mostly don't work. Not because buyers are rude — because they get dozens a week and most are identical.

The outreach that does get answered has three things in common.

1. It's specific to that shop

You can tell within two sentences whether an email was sent to forty buyers or to one. Buyers can too.

Reference something real: a brand they stock, a window they did, the fact that they carry three emerging labels and no occasion wear. One sentence is enough. It proves you looked.

2. It answers their question, not yours

Your question is "will you stock me." Theirs is "will this sell in my shop, and will you deliver."

Answer the second one. What's the retail price? What's the margin? What's the delivery window, in writing? Minimum reorder quantity? If they have to ask all four, you've already made it work.

3. It makes saying yes easy

Offer a small opening order. A trial of six to twelve pieces across a few styles beats a pitch for a full-season buy. Low risk for them, and it gets you on the rail — which is the only thing that actually sells clothes.

Then deliver on time. The single biggest predictor of a second order is whether the first one arrived when you said it would.

What kills it before it starts

No line sheet, or a line sheet without wholesale prices and delivery windows. No lookbook, or one shot on a phone against a bedroom wall. No clear brand story — buyers need to be able to describe you to a customer in one sentence.

None of this requires a big budget. A clean line sheet and decent photography on a plain background gets you further than an expensive campaign with no substance behind it.

Where production fits

Your MOQ shapes what you can offer. At 30–50 pieces per style, you can make a twelve-piece opening order across three styles and still have stock. At 500, you can't — which is why small batch production matters more for your first stockist than it does later.

Plan the capsule tight. Twelve pieces that work together sell better than thirty that don't, and the smaller number is a lot easier to fund.

For a clothing manufacturer for small US brands, the size of the market hides the problem: most factories quote American labels a minimum they can't actually sell into. A 30-piece run is the workaround, and it changes everything about the conversation.

We are a Guangzhou garment factory that has run 30 to 150 piece orders for designers in New York, LA, and smaller cities since 2016. This guide is the honest version of what small batch production USA really means — the costs, the calendar, the sizing, and the three places brands get caught.

Why a Guangzhou factory for American designers still makes sense

American small batch production exists, but it is expensive and booked months out. A domestic cut-and-sew shop at 30 pieces can quote $85 to $140 per piece before fabric.

Our small batch clothing manufacturer pricing lands lower on most styles, and the gap is not a secret: labour rates differ, and we ship 30 to 60 pieces by air in 5 to 8 days when the calendar demands it.

We are not saying local production is wrong. It is the right call for speed and for certain compliance-heavy categories. We are saying the arithmetic changes when the order is 30 pieces and the retail price is $250 to $1,000.

MOQ 30 and the small US brand: what the number really is

MOQ 30 means 30 pieces per style per colour, spread across your size run. That is a testable quantity — a capsule, a hero colour, a single silhouette you believe in.

Below 50 pieces we add a per-piece premium and print it as its own line. You should see what flexibility costs. The band where our pricing sharpens is 50 to 150 pieces per style.

If you already know your size curve from a previous season, reorder at 50 to 150. If you don't know it yet, 30 tells you more per dollar than 500 ever will.

Timeline that fits an American season

US retail seasons run early. July goods are ordered in January, and holiday product ships by September. Work backwards from the drop date, not from the quote.

Our standard clock: two to three mock-up rounds in 14 to 21 days, a final sample in the correct fabric in 28 to 42 days, then bulk production 6 to 10 weeks after fabric confirmation. Sourcing fabric adds 2 to 4 weeks.

For a first order, start the conversation 5 to 6 months before you want goods in hand. For a reorder of a proven style, 10 to 12 weeks works.

We give live order tracking and progress photos at each milestone, so the "where is my order" email never has to happen.

Sizing: the US size curve is not the Chinese one

American brands sell a different size range than the Chinese market standard, and a factory that grades for its local market will quietly drift at your extremes.

We grade from your spec, not from a house standard. Send your size chart and we build the grade table from it — including the upper sizes, where stretch fabric and standard grade rules collide.

The classic failure: the sample fits at size 8 and collapses at size 14. We check size sets at both extremes, and we fit on a live model and a mannequin at each size. If a size is wrong, we find it on the mock-up, not in the carton.

Landed cost: the number nobody quotes first

The quote is not the landed cost. American importers should add four lines before comparing anything.

• Freight. Air is 5 to 8 days and costs several times more than sea at 18 to 25 days. On 30 to 60 pieces, air usually wins.
• Duties. Most apparel enters the US with a duty rate you can look up by HTS code; plan 5 to 20% depending on the category and fabric.
• Fabric waste. 200 metres of cloth makes about 180 garments. Around 10% belongs in your number.
• Your time. Every sample round you approve late moves the delivery date. This is the cost people never add and always feel.

We itemise freight and export documents on the quote so you can see the full picture before you commit, not after.

Where 30 pieces fails

We talk people out of 30 pieces regularly, and it is the right conversation to have.

Do not order 30 of a style you have already sold at volume. The per-piece premium buys nothing. Do not order 30 when the fabric minimum forces a big buy anyway. Do not order 30 of a heavily embellished gown where setup dominates — go to 50 minimum.

Thirty is for testing: a new market, a new silhouette, a new price point. Treat it as research you can sell, and it pays for itself.

How we start with US brands

Send reference photos, measurements, and your target retail price. No tech pack? We build one — digital pattern, full size grading included. Quote within 24 hours, sampling charged separately and deducted from bulk.

For series planning, our first design tool — an AI agent we call A1, coming online soon — takes your brand positioning, style descriptions and colour theme and drafts design images to start the conversation. It is still in training, so early output is a sketch, not a final line.

If you are a small US label tired of being quoted 500 pieces you can't fund, that is exactly the call we take. Ask us how we would run your first 30.

For a clothing manufacturer for UK brands, the honest starting point is the run size: most quotes arrive at 500 pieces and die there. We run 30 to 150 per style.

Why Guangzhou for a British label? Three reasons hold up in practice: small MOQ, hand skill, and a price structure that leaves retail margin. Small batch production UK exists, but a domestic cut-and-sew shop at 30 pieces quotes £85 to £140 per garment before fabric, and the slot is booked months out. Our pricing lands lower on most styles, and the gap comes from labour rates, not from skipping inspection.

Why a British designer works with a Guangzhou clothing manufacturer

British labels that work with us fall into two groups. First collections from designers who cannot fund 500 pieces, and established brands testing a new category or price point. Both need the same thing: a run small enough to be wrong without damage.

The stereotype is that China means volume and price pressure. At 30 pieces the order is too small to be a commodity, so the work is craft — and the categories where Guangzhou workshops are strong, silk, wool and embroidery, are exactly the ones UK luxury labels buy.

We are honest about the friction: a seven-hour time difference and email, not face time. We have shipped to London, Manchester and Edinburgh since 2016, and the brands that last are the ones who accept that the first order is a conversation, not a transaction.

The UK size curve: 6 to 18 is a written spec, not a guess

British brands sell a different size range than the Chinese market standard, and a factory that grades for its own market will quietly drift at your extremes. We grade from your spec, not a house standard.

Send your size chart and we build the grade table from it, including UK 6 to 18, where stretch fabric and standard grade rules collide. The classic failure: the sample fits at a 10 and collapses at an 18. We check size sets at both extremes and fit on a live model plus a mannequin at each size, so a wrong size shows up on the mock-up, not in the carton.

Brexit changed the import maths for small batch clothing

Post-Brexit, the UK applies its own tariff schedule, and most apparel from China lands in a duty band around 12%, depending on the category and fibre. You can look up the exact rate by commodity code before you commit; we do it with every UK client.

Duties are yours, not ours — we handle export documents, packing lists and certificates of origin from the Guangzhou side. The honest part: nobody quotes the duty in their first comparison, so the landed number is usually a surprise either way. That is why we put freight and documents on the quote as their own lines, before you commit rather than after.

A timeline that fits the London season

UK retail runs early. Wholesale orders for spring are placed the autumn before, and London Fashion Week lands in February and September. Work backwards from the drop date, not from the quote.

Our standard clock: two to three mock-up rounds in 14 to 21 days, a final sample in the correct fabric in 28 to 42 days, then bulk production 6 to 10 weeks after fabric confirmation. Sourcing fabric adds 2 to 4 weeks. For a first order, start the conversation 5 to 6 months before you want goods in hand; a reorder of a proven style takes 10 to 12 weeks.

Air freight from Guangzhou to London is 5 to 8 days door to door, which is why small batches still work for UK brands that need stock in season. Sea freight at 25 to 30 days only pays off above roughly 200 pieces.

What small batch clothing costs for a UK brand

Fabric is usually the largest line, then cut-make-trim, trims, sampling and freight. Below 50 pieces we add a per-piece premium and print it as its own line, so you see exactly what the flexibility costs.

As a guide for brands retailing at £400 to £1,600, production runs roughly 22 to 30% of retail at 50 pieces. Every style differs, so we quote itemised within 24 hours of a tech pack. A single round figure is a guess, and we won't send one.

VAT, customs and the £135 rule

Import VAT into the UK is 20%, applied on the total landed value. Since 2021, goods below £135 in value are handled differently — VAT is collected at the point of sale rather than at the border, which changes how small parcel shipments are treated.

For business-to-business orders, most UK brands either import under their own VAT number or ask us to quote DDP, where we handle the lot and you receive one landed price. We have shipped both ways and will tell you which fits your accounting before you commit.

How we start with UK labels

Send reference photos, measurements and your target retail price. No tech pack? We build one — digital pattern, full size grading included — and quote within 24 hours, with sampling charged separately and deducted from bulk.

For series planning, our first design tool — an AI agent we call A1, coming online soon — takes your brand positioning, the styles you need and a colour theme, and drafts design images to start the conversation. It is still in training, so treat early output as a sketch, not a final line.

If you are a UK label tired of being quoted 500 pieces you cannot fund, that is the call we take. Ask us how we would run your first 30.

Most searches for a clothing manufacturer Middle East start with price and end with a factory that has never shipped to the Gulf. The operational side — delivery, customs, the Eid calendar — is where Gulf orders are actually won or lost.

We make 30 to 150 pieces per style in Guangzhou, and Kuwait, Saudi Arabia, the UAE, Israel and Jordan are regular destinations. This is the practical side of that work: the routes, the windows, the duties and the payment methods, so a first order arrives when the market expects it.

Why a clothing manufacturer Middle East search lands in Guangzhou

The Gulf has limited small-batch production of its own, and most volume sits in Turkey, India or China. For embellished evening wear — abayas, kaftans, gowns — Guangzhou workshops hold the hand skill and the silk and bead supply chains.

Gulf labels rarely need 500 pieces. They need 30 to 60 pieces that are right for a season, which is exactly the band a small MOQ factory runs. The region searches for a small MOQ factory in the UAE; in practice most serious small-batch runs still come from China.

Kuwait, Saudi and the UAE: what delivery really looks like

Air freight from Guangzhou to Kuwait City, Riyadh or Dubai runs 5 to 8 days door to door; sea freight to Jebel Ali, Dammam or Shuwaikh takes 18 to 25 days. On a 30 to 60 piece order, air almost always wins.

We ship air for small runs because the difference between 6 days and 20 days is usually a whole selling window in the Gulf. Export documents, packing lists and certificates of origin are prepared on our side so the shipment clears without a customs hold.

Israel and Jordan: the routes nobody quotes first

Israel and Jordan import through different lanes than the Gulf, and both are regular destinations for us. Air to Tel Aviv or Amman runs 6 to 9 days from Guangzhou; sea through Haifa or Aqaba takes roughly 25 to 30 days.

The honest part: duties differ from the GCC, and Israel and Jordan have their own tariff schedules and documentation quirks. We check the commodity code with each client before shipping so the number on the invoice matches the number at the border.

Gulf customs and the 5% that surprises

The GCC applies a common external tariff of around 5% on most apparel imports, with some categories higher. Import VAT or additional fees vary by country, and Saudi Arabia has introduced its own value-added tax layer.

None of this is huge — but it is never in the first quote, and it is exactly where first-time Gulf importers get caught. We list freight, documents and duties separately on the quote so the landed cost is visible before you commit.

One more thing to verify before you wire a deposit: that the factory has actually cleared these lanes before. We keep the customs feedback from every Gulf shipment on file and we share it — which broker, which documents, what the inspector asked for. A factory that has done the route knows the questions before they are asked.

The Eid calendar that punishes lateness

Gulf retail runs on two hard windows: Ramadan and Eid, and the bridal season. Both move every year, and both punish late delivery harder than any market we know.

Work backwards from the window, not from the quote. Sampling takes two to three mock-up rounds in 14 to 21 days, a final sample in 28 to 42 days, and bulk production 6 to 10 weeks after fabric confirmation, plus 2 to 4 weeks if we source the fabric.

We state the delivery date in the contract and hold it, because a kaftan that arrives after Eid is not late — it is unsellable.

Payment: T/T, letters of credit and the Gulf

Most Gulf clients pay by T/T, with 30 to 50% deposit and the balance before shipment. Letters of credit are accepted for larger orders and take 3 to 5 working days to confirm.

Wire transfers from Gulf banks sometimes route through correspondent banks, which can add 1 to 3 days to clearance. We confirm receipt before cutting fabric, so the clock only starts when the money is actually in hand.

We are set up for both, and we will tell you honestly which suits the order size. A letter of credit on a 30-piece run is usually overkill; on a 1,000-piece order it is normal.

MOQ 30 and the modest-fit spec

Thirty pieces per style per colour is our entry point, and for Gulf labels it buys three things: a real test of a market before committing to volume, cash that stays in your account, and the right to be wrong cheaply — a fit or colour fault costs one roll to fix, not forty.

Modest fit is part of the tech pack for Gulf clients, not an afterthought. Neckline depth, sleeve length, hemline and layering ease all get numbers, and we grade them like any other measurement. We fit on a live model plus a mannequin at each size.

How we start with Gulf and Levant labels

Send reference photos, your measurements and a target retail price. No tech pack? We build one, quote line by line within 24 hours, and keep your patterns on file so reorders don't restart.

For series planning, our first design tool — an AI agent we call A1, coming online soon — takes your brand positioning, the styles you need and a colour theme, and drafts design images to start the conversation. It is still in training, so treat early output as a sketch, not a final line.

If you are sourcing for Kuwait, Saudi, the UAE, Israel or Jordan and want a factory that has been through the Gulf season more than once, ask us how we would run your first 30.

How to find a clothing manufacturer that actually delivers is the question behind most first-order disasters. The answer is not a longer search — it is five checks before you wire a deposit.

We hear the same story every week. A brand found a factory on a directory, paid a 50% deposit on trust, and received a shipment that missed the season and the spec. The deposit was never the real risk; signing with an unverified factory was. These five checks take an afternoon, and they catch more bad matches than any contract clause ever will.

Full disclosure first: we are a Guangzhou garment manufacturer, and the factory on your list will tell you its own version of this. So run these checks on us too. They work on any supplier, anywhere.

How to find a clothing manufacturer that fits a small brand

Start with the right question. "Best clothing manufacturer" searches return directories full of trading companies and re-sellers, and most of them can quote you beautifully. The brands that succeed at MOQ 30 do not pick a factory from a ranking — they verify a shortlist, and they verify it on the ground.

Keep the list short. Three factories, not thirty. Then spend the afternoon on the five checks below, in order. If a factory fails one of them, it saves you the other four.

Check 1: The licence and an address you can verify

Ask for the business licence and the registered address before anything else. A real factory answers with documents in the legal company name — ours is Guangzhou Zitong Fashion Co., Ltd. A trading company can still be a fine partner, but you should know which one you are paying, because the margin structure is different.

Verify the address against a map, then ask for a live video walkthrough of the floor. We do these regularly: camera down the cutting table, across the sewing lines, into the finishing room. If a supplier cannot show you the floor for five minutes, they do not have one.

Check 2: A line-by-line quote, not a single number

A one-line price is a red flag by itself. A $18 per piece figure can hide fabric substitution, missing trims, or sampling charged twice. Ask for fabric, cut-make-trim, trims, sampling and freight as separate lines — the same structure we quote within 24 hours of a tech pack.

Line-by-line quotes are how you compare suppliers honestly, and how you see the fixed costs that make MOQ 30 expensive. When setup appears as its own line, you can decide whether 30 pieces is the right call or whether 50 makes the maths cleaner.

Check 3: The QC standard, on paper

Ask what inspection standard the factory runs, and get it in writing. Most volume factories quote AQL 2.5; we run AQL 1.0 with size-by-size measurement, which allows roughly half the defects. On a run retailing at $500 and up, that gap is thousands of dollars in returns.

Then ask the harder question: do you inspect every piece or a sample? We inspect three times — incoming fabric, inline, and finishing. A factory that spot-checks a 30-piece order is checking almost nothing at all. The number matters more than it sounds; we break down why in AQL 1.0 vs 2.5: what the number actually controls.

Check 4: How they sample

Sampling is where a factory shows its real level, because nothing is hidden yet. Ask how many mock-up rounds are included, what fabric they mock up in, and whether the final sample uses your actual fabric. Then ask to keep a sealed sample — the signed, dated garment both sides keep as the reference for the whole run.

Two to three mock-ups in 14 to 21 days is normal for a first run. If a factory promises one round and a ship date in four weeks, they are selling you speed they do not have. Our mock-up process is slower on purpose, and the fit is better for it.

Check 5: Calendar honesty

Ask when the delivery date becomes real. The honest answer: after fabric is confirmed, not before. Any factory that books a line before the mill confirms fabric is promising a date they do not control — and that is the exact failure behind most late orders.

Our sequence is fixed: mock-ups in 14 to 21 days, a final sample in 28 to 42, bulk production 6 to 10 weeks after fabric confirmation, plus 2 to 4 weeks if we source the fabric. If your season cannot absorb that, you need to know in week one, not week nine. This is the same logic as our lead time breakdown.

Where to look first: directories, shows and referrals

Directories are the weakest source and the most used. Trade shows — Canton Fair in Guangzhou, Première Vision in Paris — put you in front of the floor, which is worth a month of email. Referrals from brands in your own price band are the strongest of all, because someone has already run the checks for you.

Whichever channel you find a factory through, the five checks still apply. The channel decides who is on your shortlist; the checks decide who stays on it.

What the five checks catch

Run in order, they filter most of the industry's failure modes before money moves: fake factories, hidden margins, invisible QC, rushed sampling and fantasy calendars. We lose enquiries to these checks sometimes — a brand finds a cheaper quote elsewhere and we never see them again. We also see them again six months later, on our case studies page, after the cheap quote did what cheap quotes do.

Finding a clothing manufacturer is not a search problem. It is a verification problem, and five checks are enough to do it properly.

Clothing manufacturer for US designers: the phrase usually comes with one worry — can we work with a Guangzhou factory without ever visiting? Yes, and most of our American clients have never stepped into the workshop. The workflow is built for distance.

Half of our clients are in the US, from New York to LA, and they run first seasons remotely because a visit is not required. What is required is structure: a weekly rhythm, video at every gate, and a sealed sample you approve in writing. Here is how that actually runs.

Working with a clothing manufacturer in Guangzhou from the US

The first difference from a local factory: your project moves in weekly batches instead of daily chatter. Monday you send feedback, Tuesday we pattern, Friday you get mock-up photos and video. One focused round trip beats ten scattered messages, and the time gap works in your favour — you wake up to progress, not to questions.

Every milestone is documented the same way: photos of your actual fabric on the cutting table, video of the first piece off the line, a filmed final inspection before the cartons close. You see the run as it happens, which is more than most in-person visits ever capture.

How the remote workflow actually runs

Step one is a brief, and it does not need to be a tech pack. Reference photos, your measurements, target retail price — we build the tech pack, digital pattern and full grading. Step two is two to three mock-up rounds in 14 to 21 days, reviewed by video and photo. Step three is the final sample in your actual fabric, sealed and signed, shipped to you or kept on file.

Step four is bulk, 6 to 10 weeks after fabric confirmation, with three inspections per garment. Step five is freight: air in 5 to 8 days or sea in 25 to 30. On a 30 to 60 piece order, air almost always wins, because the extra cost buys a whole selling window.

The 12-hour gap and why it helps

New York is 12 hours behind Guangzhou, LA is 15. At first that feels like a delay; in practice it is a handoff. You send notes at the end of your day, the workshop works through its day, and answers are waiting when you open the laptop. Decisions that take a week in a shared office take two days here, because nobody is idle — someone is always working.

The rule that keeps it honest: every comment is answered within 24 hours, in writing, with photos if it touches the garment. If a factory cannot commit to that, remote production will frustrate you no matter how good the sewing is.

What replaces a factory visit

We run live video walkthroughs on request — down the cutting table, across the lines, into finishing. For the visit you would normally make at first-piece, you get video of the first garment measured against the sealed sample. For the final audit, you get the filmed carton-by-carton inspection we send before every shipment.

None of this replaces being there. But for a 30-piece first order, it replaces 90% of the value of a visit at 1% of the cost — and the saved flight money usually funds the sampling round.

The US maths: duties, freight and the season window

Import duties into the US run roughly 5 to 20% depending on the HTS category, and the entry is on your side; we handle export documents, packing lists and certificates of origin. Freight for a small batch runs $3 to $6 per piece by air. Factor both and the landed cost still beats most small batch production USA options by 30 to 50% at the same quality.

Work backwards from your drop date. For a first order, start 5 to 6 months before you want goods in hand; reorders with patterns on file compress to 10 to 12 weeks. The calendar, not the quote, is the real constraint — the same logic we lay out in our lead time breakdown.

What to send in your first email

Send reference photos, a size chart, your target retail price, and the season you are aiming at. We quote line by line within 24 hours and tell you honestly if the silhouette fights the fabric or the price point.

For series planning, our first design agent — A1, coming online soon — takes your brand positioning, the styles you need and a colour theme, and drafts design images so you can brief us with visuals instead of words. It is still in training, so treat its early output as a sketch; we take it from there.

If you are a US designer testing a first run, ask us how the workflow would look for your specific silhouette. The answer is usually the same: send the brief, watch the photos, approve the sample, and the goods arrive in time for the season.

Choosing an abaya manufacturer for Gulf brands comes down to three things before price: the modest fit spec, the fabric, and whether MOQ 30 is a real number or a sales line. We are a Guangzhou garment factory that has run 30-piece abaya and kaftan orders for Kuwait, Saudi and UAE brands since 2016. Most first emails we receive ask one question — can you hold Gulf quality at small volume? Yes, and here is the spec that makes it work.

Send us the lengths you need, the fabric you want and your delivery month, and we quote within 24 hours. The longer version is below, because the details — how a hem falls over a hijab-friendly layer, whether a shoulder seam sits right under a shayla — are what separate a factory that understands modest wear from one that treats it as a long dress.

Abaya manufacturer for Gulf brands: why the fit spec comes first

Most abaya problems start before the fabric is cut, in a pattern graded like Western outerwear. Gulf abayas are bought on length, coverage and fall, not on bust fit. So we put three numbers in every tech pack: front length, back length and sleeve length, measured on the finished garment against your size chart.

Kuwaiti buyers usually want a longer, fuller line; Saudi clients often ask for cleaner tailoring around the shoulder. We do not guess which — we grade from your numbers, and we check both size extremes on a size set, because hem fall is exactly where small-batch grading drifts.

One mistake we admit making early on: shortening the first mock-up to save fabric. That is how length errors happen. Now the first mock-up is always cut at full length, and the hem is reviewed on a live model or mannequin at the height you specify before bulk.

Nida, crepe or silk: what Gulf abaya fabric really needs

For everyday abayas, nida — a matte polyester microfiber — dominates the market for a reason: it drapes, resists creasing and holds colour after years of washing. For evening and occasion pieces, our clients move to silk crepe or crepe de chine, usually 16 to 19 momme, where the matte finish reads expensive and the weight gives the skirt a clean fall.

Two properties decide whether a fabric looks cheap, and both are testable: sheen and weight. A glossy poly that shines under boutique light photographs like costume; nida and silk crepe are chosen partly because they stay matte. We verify weight by grams per metre and fibre by burn test at intake, and we wash-test every new abaya fabric for shrinkage before cutting — a crepe that shrinks 4% after the first wash will pull a hem up an inch.

For summer, breathability matters as much as drape. We will tell you honestly when a beautiful heavy fabric is wrong for a June delivery to Riyadh.

Embroidery and stonework that survive Gulf wear

Abaya and kaftan embroidery is usually the make-or-break detail, and the failure mode is specific: motifs that shift shade between rolls, or stones and sequins that shed after a season. We digitise every motif for the actual fabric, seal the thread colour against it, and anchor stones with a double-back stitch, pull-tested at 8N on 10% of the run.

Machine work at 12 to 14 stitches per centimetre is right for most runs of 30 or more; hand accents — a crystal cuff, a hand-finished hem — go in where the retail price supports them. If a motif will not hold at small volume, we say so before sampling, not after. More on the shade side is in our colour matching guide.

MOQ 30 abaya production and the Eid calendar

MOQ 30 is per style per colour, which is how Gulf boutiques test two or three silhouettes without betting the season. Two styles in compatible fabrics can share one production slot, and that is the trick most new clients miss — a capsule of three styles at 30 each is often cheaper than one style at 90, because setup is spread across the same line time.

The real constraint is the calendar, not the unit price. Sampling runs two to three mock-up rounds in 14 to 21 days, a final sample lands in 28 to 42, and bulk takes 6 to 10 weeks after fabric confirmation, with air freight 5 to 8 days to Kuwait, Saudi or the UAE. For an Eid drop, start four to five months out; a Ramadan capacity is booked early, and an abaya that arrives after Eid is not late — it is unsellable.

What to send an abaya manufacturer in your first email

Reference photos of the drape you want, a length table for front and back, the fabric (or ask us to source), trim preferences and the retail price you are aiming at. We quote line by line and flag when a silhouette fights the fabric or the price point.

For a whole capsule, our first design agent — A1, coming online soon — drafts design images from your brand positioning, the styles you need and a colour theme, so you can brief us with visuals before the first pattern. It is still in training, so treat its early output as a sketch; we take it from there.

Every Gulf order we ship is inspected three times per garment against the sealed sample, and the shipment is video-documented carton by carton. That is the short version of how MOQ 30 modest wear is supposed to work.

UK import duty on clothing from China is 12% for most garments under the UK Global Tariff, and import VAT adds 20% on top of the duty and the freight. We are a Guangzhou garment factory that ships small batches to London and Manchester every month, and UK brands are usually surprised by the border bill — not by its size, but by how predictable it is when someone writes the numbers down.

What does a 30-piece order actually cost at the border? Take the ex-works price, add roughly 12% duty, then add 20% VAT on the goods plus duty plus freight. Stacked, that is about 34% on top of the factory price before shipping — and most of it is reclaimable if you are VAT registered. Here are the three numbers that decide the real bill.

UK import duty on clothing from China: the 12% default

Most apparel sits in chapters 61 and 62 of the tariff, and under the UK Global Tariff most of it carries 12% duty. Some categories differ — certain silks and knitwear fall lower, a few synthetics run higher — which is why the commodity code for each style matters more than a rule of thumb.

China has not had developing-country preference on this route since 2021, so the full rate applies; there is no workaround in how we invoice, and we would not pretend otherwise. What we do is check the commodity code for every style before shipping and write it into the export documents, so the number your customs broker quotes matches the goods in the carton.

UK VAT at 20%: who pays and when it hurts

Import VAT is charged on the goods, the duty and the freight together — that is the stacking that surprises people. A VAT-registered UK business reclaims it on the next return, so the real cost is the cashflow gap, not the tax. A non-registered startup or a private buyer pays it and cannot reclaim.

The old low-value relief below £15 disappeared in 2021, so there is no small-consignment exemption to plan around — VAT applies from the first carton. We flag this before you commit because it changes the retail maths on a first 30-piece order.

What freight adds per piece

Air freight from Guangzhou to the UK runs 5 to 8 days and adds roughly $3 to $6 per piece on a small batch, depending on weight and volume. Sea takes 25 to 30 days and only pays off above roughly 200 pieces. On a 30-piece order, air wins, because the extra cost buys a selling window — and the duty and VAT are charged on a lower freight figure than you might expect, since we ship at commercial weight, not inflated volumetric guesses.

A worked example: what a 30-piece order actually pays

Take a silk evening dress quoted ex-works at $45 per piece, FOB Guangzhou. Thirty pieces is $1,350. Duty at 12% adds $162. Air freight at $4 per piece adds $120. Import VAT at 20% is charged on the goods plus the duty plus the freight — $1,350 plus $162 plus $120 is $1,632, and 20% of that is $326.

The full landed bill is about $1,958, or $65 per dress before any UK margin. If you are VAT registered, the $326 comes back on your next return, which brings the real cost to about $54 per dress. If you are not registered, the $326 is part of your true cost, so it belongs in the retail price before you commit — not discovered after the invoice lands.

Two details we always write down in the quote: the duty is not reclaimable, and it is calculated on the commodity code we verify against the UK tariff for that exact style. When a client wants a fixed landed number instead of managing the border themselves, we can quote DDP through a UK customs partner — the price includes duty and VAT, and there are no surprises at delivery.

Documents that stop customs delays

Border holds on small shipments are almost always paperwork problems: a commercial invoice that does not match the packing list, a missing certificate of origin, or a commodity code that describes a dress as a blanket. We prepare the export side — commercial invoice, packing list, certificate of origin and verified commodity codes — before the cartons close.

One honest caveat: we are a factory, not a customs broker. For complicated entries, use one; for a standard 30-piece garment shipment, our documents plus your broker's code check is normally enough. That combination is why our UK clients rarely see a hold longer than a day.

Why the maths still favours Guangzhou for a small run

Add the duty, the VAT and the air freight, and a 30-piece order from Guangzhou still lands well below UK small-batch production at the same quality — for embellished silk and evening work, the saving is usually 30% or more. The trade-off is lead time and planning, which is why we ask UK clients to start first orders four to five months ahead of the season, as set out in our lead time breakdown and the UK brand guide.

The border is not where small-batch sourcing fails. It fails when the numbers are hidden until the invoice arrives. Ask for the commodity code, the duty rate and the VAT treatment in writing before you commit, and the bill at the border becomes a line you planned for, not a surprise.

Choosing a kaftan manufacturer for Gulf brands usually starts with fabric and fit. It should start with embroidery. In Kuwait, Saudi Arabia and the UAE, the ornament is what makes a kaftan worth AED 1,400 — or what makes it look like a bathrobe with beads.

Buyers ask us one question first: will the embroidery look this good after ten wears and a Gulf summer? The honest answer is that it depends on decisions made before the first motif is stitched — the technique, the backing, the thread, and the anchor behind every stone. We are a Guangzhou garment factory that has run kaftan and abaya ornament work since 2016, and this guide is how we make those decisions at MOQ 30.

Kaftan manufacturer for Gulf brands: why the ornament decides the sale

A Gulf customer does not buy a kaftan for its seams. She buys the collar of embroidery, the beaded cuff, the motif that photographs well and survives the evening. Ornament can be 40 to 60 percent of what a decorated kaftan costs to make, which means the factory's embroidery floor is your real supply chain — not the cutting table.

That is why we check three things before we quote: how the motif is placed on the silhouette, how it is stitched onto the fabric, and how it is anchored so it does not shed. Silhouette comes first because a kaftan is cut floor-length with a clean fall, usually fully lined, with sleeves and side slits that let the fabric move. Placement that ignores the drape will pucker on the body even if the embroidery itself is perfect.

We made that mistake early on. We once sealed a motif on plain cotton sampling cloth because it was cheaper than the real silk. On the silk, the same motif sat stiff and the puckering showed in every photo. Now every placement is sealed on the actual fabric, never on a substitute.

Hand or machine embroidery: how a kaftan manufacturer should choose

Machine embroidery is the right call for most 30-piece runs. It is repeatable, which matters when every garment must match the sealed sample, and at 12 to 14 stitches per centimetre the work reads as premium on silk and viscose grounds. Digitising a motif costs roughly USD 30 to 80 once per design, and each stitched garment adds USD 6 to 18 depending on the area covered.

Hand work earns its place on accents: a crystal cuff, a couched cord edge, a hem finish that machines cannot copy. It adds roughly USD 25 to 120 per garment on a kaftan, and the cost is stable only if the motif area is fixed in the quote. If a client wants hand accents at MOQ 30, we say yes — the ornament spec just has to be precise, because hand work drifts when the brief is vague.

Ask the factory which technique it is quoting. A price that does not name hand or machine is a price that will change at the sample stage.

Beads and stones that survive Gulf wear

The failure mode Gulf buyers fear most is shedding: stones lost after a season, sequins that dull, thread that snaps under the arm. We anchor every stone with a double-back stitch and pull-test at 8N on 10 percent of the run. Bead weight is checked per square metre so heavy crystal does not stretch a silk ground, and thread colour is sealed against the final fabric under daylight and boutique light.

Fabric and embellishment fail together, which is why we test shade at intake before a single motif is stitched — the same discipline our colour matching guide describes for dye lots. If a motif will not hold on the fabric you chose, we flag it before sampling, not after.

Locking the ornament in the mock-up rounds

Embellished kaftans need their own sampling rhythm. Round one fits the silhouette in muslin. Round two brings the real fabric with the motif stitched in place and a density sample of the beadwork — this is the round that decides everything. Round three is for heavy beading or a second colourway, and it is optional.

Two to three rounds land in 14 to 21 days, and the sealed sample follows by day 42. The seal matters more on decorated work than on plain styles: the signed sample is the reference for bead size, spacing, stitch density and colour, and bulk is measured against it piece by piece. For more on why rounds go wrong, our mock-up guide covers the failure points.

Planning the Gulf calendar: weddings, Ramadan, Eid

Embellished kaftans sell on the occasion calendar, not on a retail season. Winter weddings run from roughly October to February across the Gulf, and Ramadan and Eid shift about 11 days earlier every year, so the buying window never sits still.

We plan beaded drops backwards from the event: two to three mock-up rounds, a sealed sample by week six, then 6 to 10 weeks of bulk after fabric confirmation, plus 5 to 8 days of air freight to Kuwait, Saudi or the UAE. For a Ramadan or Eid drop, start five months out; ornament capacity books early, and a decorated kaftan that lands after the occasion is not late — it is unsellable. Our Gulf timelines guide shows the same maths for plain runs.

What to send a kaftan manufacturer in your first email

Send a reference photo of the drape and ornament you want, a sketch of motif placement, the retail price you are aiming at, and the delivery month. That is enough for a line-by-line quote within 24 hours, including the hand-versus-machine decision with its costs.

For a whole capsule, our first design agent — A1, coming online soon — drafts design images from your brand positioning, the styles you need and a colour theme, so placement arrives as a visual brief instead of a guess. It is still in training, so treat early output as a sketch; we take it from there.

Every decorated order we ship is inspected three times per garment against the sealed sample, and the shipment is video-documented carton by carton. The full sequence is on our production process page, common fit and beading questions are answered on the FAQ page, and you can read decorated kaftan results on our case studies page.

Comparing small batch production USA against a Guangzhou factory usually stops at one number: the unit price. On paper the Chinese quote looks unbeatable. In practice the honest gap is smaller — and sometimes it flips. Here is the comparison we give American designers who ask whether offshore still makes sense at 30 pieces.

The short answer: for sewn luxury in small runs, Guangzhou typically lands 30 to 50 percent below US domestic at the door, after freight and duty — when the order is planned rather than rushed. Domestic wins on speed and near-zero language risk. The real question is which of those you are actually buying, and we are honest when the answer is domestic.

Small batch production USA: what domestic really charges

A US small-batch studio in New York or Los Angeles quotes labour that reflects its rent. Cut-and-sew on a lined dress at 30 to 60 pieces commonly runs USD 45 to 90 per garment, plus pattern and grading at USD 400 to 900 per style, plus sampling at USD 500 to 1,500. Those numbers vary by studio, but they explain why a 30-piece capsule can carry five figures of development cost before a single unit sells.

What domestic buys is speed and proximity. A restock can turn in two to three weeks. A fit issue is solved with a phone call and a same-day courier. If your brand sells on a made-in-USA story or needs a fast second drop, that premium is a marketing cost, not waste.

Small batch production in Guangzhou: the same order, itemised

At a Guangzhou factory the same lined dress at MOQ 30 typically runs USD 12 to 30 in cut-and-sew labour, with fabric sourced 15 to 40 percent below US wholesale for silks and wools. Sampling costs USD 150 to 400 per style and is credited against bulk on confirmation. Freight adds USD 3 to 6 per piece by air, landing in five to eight days.

Duty is the line Americans forget. Most dresses sit in US tariff chapter 62 at roughly 16 percent, and additional tariffs on goods from China have been added and changed several times in recent years — so the honest instruction is to verify the current rate for your exact commodity code with a broker before you commit. We write the code into the export documents so the number you are quoted matches the goods in the carton.

Worked example: a silk dress at USD 32 ex-works Guangzhou. Add air freight of USD 4 and duty of roughly 16 percent, and the landed cost is about USD 42 per piece. The same dress from a US small-batch studio often lands at USD 70 to 95. The saving is real — and it is why our American clients order at MOQ 30 rather than 500 when they are testing a style.

Where the gap shrinks, and where it flips

Domestic wins the moment speed beats price. A reorder needed inside three weeks, a showroom sample rushed for a buyer meeting, a five-piece capsule for a brand launch — all should probably stay in the USA. Offshore also loses its edge when the design needs daily hands-on iteration, which is why our remote workflow guide exists: we structure the process so most US clients never need that.

Cash protection is the quieter argument for Guangzhou. A 500-piece domestic minimum is a five-figure bet before the first dress sells; the same style at MOQ 30 tests the market for a fraction of that cash. For a young American label that is often the difference between running a pilot and running a gamble.

China wins on decorated and hand-finished work, where US small-batch capacity is thin. Beading, embroidery and silk evening pieces at small volume are exactly the orders where the 30 to 50 percent gap holds, because the craft labour is scarce and expensive in American studios. We also admit the risk: tariffs have moved without warning, and a change between quote and shipment can shift the math — which is why we re-confirm the commodity code and duty treatment in writing at confirmation, not at the border.

How a US brand tests Guangzhou without betting the line

Test with one style at MOQ 30, not a whole collection. You get two to three mock-up rounds in 14 to 21 days, a sealed sample by week six, then 6 to 10 weeks of bulk. Video evidence arrives at every gate — fabric intake, inline inspection, final check — so the pilot is verifiable from across the Pacific.

Most of our US clients have never visited the workshop. They judge us on the pilot: the sample quality, the node dates, the inspection report with numbers. Our US small brand guide covers what to send in the first email, and the production process page shows the full sequence before you commit a deposit.

The honest bottom line

Choose Guangzhou when the order is planned, the craft is the point, and the cash risk of a 500-piece minimum would hurt. Choose American production when the deadline is days away or the made-in-USA story is your price.

We would rather you make the right call than take the cheapest one — that is why this comparison exists. If the math works for your style, our cases page shows what remote US clients actually ship; if it does not, we will tell you before you pay for a sample. Freight and duty questions are answered on the FAQ page.

Ask any small MOQ production factory what its floor can do at 30 pieces and the brochure answers fast. The honest answer is a list: five techniques we run happily at MOQ 30, four that quietly triple the price, and the checks that keep the difference off your invoice.

The short answer: cut-and-sew in silk, wool, cupro and linen, machine embroidery, digital and screen print, and hand finishing all run well at 30 pieces. Custom jacquard, gradient prints and anything that needs a mill to spin a new base usually do not. The rest of this guide is how we sort one from the other before we quote — not after you have paid.

Small MOQ production: five techniques that work at 30 pieces

Cut-and-sew from stock cloth is the backbone. A lined dress, blazer or shirt in a fabric the mill already runs costs about the same to make whether the order is 30 or 300; only the setup spread changes. That is why most of our 30-piece runs are wovens and knits in silk, wool, cupro and linen.

Machine embroidery is the second. A programme costs USD 30 to 80 once per design, then USD 6 to 18 per garment depending on the area covered, and repeatability is exactly what a 30-piece order needs. Hand accents — a crystal cuff, a couched cord edge — still work at 30, quoted at USD 25 to 120 per garment, as long as the motif area is fixed in the quote.

One honest note on sampling: at these volumes, development is proportionally larger than the bulk. Budget two to three mock-up rounds in 14 to 21 days and a sealed sample before cutting starts, and the 30-piece run itself becomes the predictable part of the calendar.

Digital and screen print survive small runs too. Digital charges a one-off set-up per design; screen print needs 50 to 100 metres per colourway, which a 30-piece style usually reaches once the yardage is added up. What dies at 30 is rarely the printing — it is the fabric the print needs.

What a 30-piece run cannot do cheaply

Custom jacquard needs a mill to set a loom, and mills ask for 800 to 1,500 metres before they start. Custom-dyed stretch wools and cupro behave the same way, usually from 600 metres up. A gradient or engineered print that must match across panels carries the same logic.

Trims hide the second minimum. Custom metal buttons start around 1,000 pieces, printed labels around 500, and a bespoke woven label rarely makes sense below 300. We flag these before sampling, and we have steered clients to stock cloth and printed main labels when the maths did not work.

We have also turned work down. If a client needs a custom-woven jacquard at 30 pieces, the honest answer is a quote for the mill minimum and a much bigger order than they planned — not a promise we cannot keep.

How a low MOQ factory keeps 30 pieces affordable

Three habits do the work. We cut one marker for the whole run instead of one per size, we plan each capsule so its styles share fabric and trims, and we run small orders through multi-skill operators rather than a fixed production line.

A capsule marker alone usually saves 8 to 12 percent on fabric, because sizes nest together on one spread. Batching two styles cut from the same cloth in a single cutting session saves again, which is why we ask to see the whole capsule before we quote, not style by style.

Below 50 pieces the fixed lines — pattern, grading, marker, setup — are spread thin, and that is the per-piece premium on the quote. We print it as its own line so you can see exactly what 30 costs versus 150.

Three rules that protect a 30-piece budget

First, seal every finish on the real fabric, never on substitute cloth. The mock-up is where placement, puckering and drape are decided, and our mock-up guide explains why that round carries the whole project.

Second, confirm the stock-cloth lot before you approve the sample. Two lots of the same silk can differ in shade and width, and the checks in our fabric sourcing guide catch that before cutting — not after.

Third, narrow the capsule before you spend on sampling. Three to eight styles that share cloth and trims cost a fraction of eight unrelated ones, and one tight range sells better than a loose one. When a founder is deciding which styles deserve a run at all, our first design agent — A1, still in training and coming online — drafts design images from your brand positioning, the styles you describe and a colour theme, so the shortlist arrives as a visual brief instead of a guess.

Treat A1's early output as a sketch. The production decision still runs through sealed mock-ups and a signed sample — the agent shortlists, the floor decides what is sewable.

When 30 pieces is the wrong number

Thirty is for testing a style, a stockist meeting, or a first season where you can absorb a miss. Once a style sells through at full price and the channel asks for depth, order 60 to 150 — the per-piece price drops at those tiers, and the small-run premium stops applying.

We say this out loud because it protects both sides. The arithmetic of 30 versus 500 is on our MOQ cost guide, and the honest rule is simple: run small when the answer is unknown, run bigger when the answer is already on the books.

Every order we ship is inspected three times per garment against the sealed sample. The full sequence is on our production process page, fit and sampling questions are answered on the FAQ page, and past 30-piece runs across four markets are documented on our case studies page.

Most garment manufacturer search results are trading agents with a desk, a phone and somebody else's factory photos. Before you compare a single price, work out which of the four models you are actually emailing: an agent, a CMT workshop, a full-package factory, or a volume mill. The model decides who owns your pattern, who buys the fabric, and who answers when the sample drifts.

For an independent brand ordering 30 to 150 pieces with real design input, a full-package factory with its own sewing floor is usually the right fit. It is the only model where one party owns pattern, fabric, trims and quality control from the first sketch to the sealed carton.

The four models hiding behind the phrase garment manufacturer

A trading agent owns nothing. It forwards your files to a factory it may never have visited, adds 15 to 30 percent, and steps out of the way when something goes wrong. Some agents are excellent operators; the risk is that you cannot verify which kind you have.

A CMT workshop cuts, makes and trims, and expects you to supply the fabric. You hold the mill minimums, the dye-lot risk and the shipping — control is high, but so is the workload, which suits brands that already buy cloth.

Read the CMT quote carefully, because labour is only the visible line. A lined silk blouse might read USD 12 to 30 a piece for sewing alone, with the fabric bill separate and yours. If the mill minimum is 600 metres and your capsule uses 120, you now own five seasons of cloth — planned, that is an asset; discovered late, it is a warehouse bill.

A full-package factory sources fabric and trims, grades and samples, and runs the sewing on its own floor. One quote, one owner, one inspection standard — this is the model most of our clients work with, and the one built for a 30 to 150-piece order.

A volume mill lives at 300 to 1,000 pieces and up. The unit price is low and the minimum is the catch: 40 pieces is not a production run to a floor that ships 100,000 pieces a month.

What changes when the model is wrong

Think full package and discover CMT, and the fabric bill lands on you — including the 600 to 1,500-metre minimums we describe in our small MOQ production guide. Think agent and find a forwarding desk, and your factory can substitute cloth or miss a season with nobody accountable.

The quiet cost is quality control. A volume mill defaults to AQL 2.5 because it cannot afford to slow the line, and the difference matters most on decorated work — that is why our AQL guide exists.

We have rebuilt projects where the model was wrong twice over: an agent who routed the sewing to a third-country mill, and a CMT arrangement where the client-supplied fabric arrived a different shade. Each rebuild started by naming who owned what — then fixing the process around it.

How a small brand should choose

If you already buy fabric, own the colour story and know the mill minimums, CMT can save money. If you are designing from a sketch and need one accountable partner, choose full package — you trade a little control for a lot of line-of-sight.

Four questions reveal the model in one call. Who owns the patterns? Where is the sample sewn — can we see the floor on video? Who issues the fabric invoice? Who holds the trims? An agent answers all four vaguely; a factory answers with documents.

The deposit checks in our find a manufacturer guide still apply on top: licence, a line-by-line quote, QC on paper, sampling terms and a written calendar.

A worked example: the same 40-piece blazer under two models

Take a lined wool blazer at 40 pieces. Under full package the quote arrives as one page: fabric at a set price per metre with the quantity stated, pattern and grading, two mock-up rounds, trims, sewing, pressing and packing, plus sampling credited at confirmation. You can see the whole cost of the style before you commit.

Under CMT the same blazer looks cheaper at first — sewing only — until you add pattern and grading at USD 300 to 900 per style, a sampling charge with no credit unless you negotiate one, your own fabric purchase against mill minimums, and a freight booking you arrange and insure yourself. Add a 10 to 15 percent buffer for the things that shift, and the two models often land within touching distance.

Neither is universally cheaper. Full package costs more per line but fewer surprises; CMT rewards brands that already run a supply chain. The mistake is comparing the CMT labour line with the full-package total and concluding one factory is greedy.

What we run, and what we tell you straight

We are a full-package clothing manufacturer with our own Guangzhou floor — about 40 people, running since 2016, at 30 to 150 pieces per style in silk, wool and embroidery. When a client genuinely needs CMT because they already hold the fabric, we say so and quote it honestly.

Garment manufacturer types do not decide quality; accountability does. We would rather lose a job to the right model than win it with the wrong one — our process page shows the full sequence, common factory-model questions are answered on the FAQ page, and runs under each model are documented on our case studies page.

A production quote usually arrives after the design is finished. Ours often starts the other way round: you name the price you need to hit, and we plan the garment backwards, choosing fabric, lining, construction and trims against your target before a single sample is cut. Budget-based production is how a growing share of our quotes now begin, because it answers the only question that matters early — is this garment buildable at your number?

So yes: tell us your target price per piece and we work from there. Send a sketch, a reference garment or a rough spec with a number attached, and we come back with options — what the price buys in silk, what it buys in cupro, and where the design itself has to change to stay honest. What we will not do is quote one quality and quietly deliver another.

Step 1 — Decide which number: FOB price or landed cost

First, name which number you mean. FOB at our Guangzhou door is not the same as landed in New York or London: freight, duty and VAT sit between them, and a USD 100 landed target usually leaves only USD 85 to 90 for the factory once those are stripped out. We plan against the number that protects your margin, and we write the border into the quote so the target cannot be quoted against the wrong one.

Second, say whether the target is fixed or a range. A USD 60 to 70 range lets us return two honest specifications — one at the bottom of the range, one at the top — instead of one stretched plan. A fixed number is fine too; it just shortens the fabric list faster.

Step 2 — Choose the fabric weight the target price can carry

Fabric is 30 to 45 percent of most small-batch quotes, so it is the first lever we pull. A USD 60 dress target points to a silk twill or cupro at 12 to 16 mm rather than a heavier charmeuse; at USD 90 we can keep the charmeuse and add a lining. Two cloths that look alike can sit USD 4 to 8 per metre apart, and a dress needs 1.3 to 1.6 metres of body fabric before the lining is counted.

We quote against a named mill and a named weight, never the word "silk" on its own. That keeps the number from drifting when the cloth is bought, and it is the same discipline described in our fabric sourcing guide. If a lot runs heavier than planned, the difference appears in the quote before cutting — not in a surprise on the invoice.

Step 3 — Spend the construction budget where customers look

A blazer at USD 110 and a blazer at USD 160 can share the same wool. The gap is structure: a half-canvas front, a shaped sleeve head, under-collar work and lined pockets. We spend where a customer looks and feels — lapel roll, shoulder line, pocket mouth — and save where they never look.

Dresses and jumpsuits follow the same rule. An invisible zip and a bound seam earn their cost; an extra internal frill does not. Tell us which details your customer touches, and we protect those lines first and trim the rest.

Step 4 — Count trims before they eat the unit price

Trims carry fixed costs that a 30-piece run cannot spread. Custom metal buttons start around 1,000 pieces, a bespoke woven label rarely makes sense below 300, and printed main labels cost a fraction of both. When a target price is tight, we move those decisions to stock options and write the saving on the quote.

Zips, elastics and hanging loops look small on a spec and add up fast. The only quote we trust names every trim line by line, which is why our cost per garment guide exists. If a line is missing, the cost has not gone away — it has just moved somewhere you will not see until later.

Step 5 — Lock the spec, the price and the calendar in one pass

Once the choices are made, we write them into one document: fabric with mill and weight, every trim, the construction notes, the FOB price, and dates for mock-up, sealing, bulk and shipment. Nothing moves after that without a written change order, because a drifting spec is how target prices quietly become overruns.

Sampling at MOQ 30 is a real cost — usually USD 150 to 400 per style, credited when you confirm. Two focused mock-up rounds beat five drifting ones, and our mock-up guide explains why that first round decides the whole project.

Step 6 — If the number will not work, we say so before sampling

Sometimes a target is not reachable at the quality a garment needs, and we say so before sampling. A fully canvassed wool blazer at USD 55 FOB is not something we will promise, and neither is a beaded silk gown at USD 40. We show the arithmetic — the cloth alone at the right weight costs more — and offer the next honest option: a lighter fabric, a simpler construction, or a corrected number.

That honesty costs us a few jobs and keeps the ones that matter. A brand that knows its real floor price can plan a whole season around it, and that is exactly what budget-based production is for.

When the design itself is still loose, the method works one step earlier. Our first designer agent, A1, is in training and coming online soon: you give it your brand positioning, the silhouettes you want and a colour story, and it returns design drawings, so a budget conversation can start from a visual brief instead of a guess. The agent shortlists; our floor still decides what is sewable through sealed mock-ups.

The full sequence behind a price-locked order — mock-up rounds, a sealed sample, three inspections per garment and export packing — is documented on our production process page, and the budget, deposit and fit questions buyers ask most are answered on our FAQ page. How MOQ 30 keeps unit costs down is explained on our low MOQ landing page, and orders we planned this way, from silk dresses to wool blazers, are on our case studies page.

A blazer and a dress can share the same MOQ, the same factory and a similar fabric bill, and the blazer will still quote 30 to 60 percent higher. That gap is not a markup; it is construction. Tailoring asks a sewing floor to do work a dress never needs, and the difference shows up in almost every line of the cost — pattern, cutting, sewing minutes and pressing.

Brands ask us this every week when a capsule contains both. The short answer: a simple dress is cut from six to ten pattern pieces, a tailored blazer from twenty-five to thirty-five, and every extra piece carries sewing, pressing and inspection time that a 30-piece run cannot spread thin. Here is where the money actually goes.

Pattern pieces and cutting: why the blazer starts behind

A sheath dress can be six to ten pattern pieces; a structured blazer with a two-piece sleeve, an under-collar and welt pockets runs twenty-five to thirty-five. Each piece must be cut, matched, sewn and pressed, and at 30 pieces that work lands on a short run, which pushes the per-piece cost up before a single seam is sewn.

Cutting tells the same story. A blazer consumes more cloth per garment than a dress of the same length because the pieces do not nest as tightly, and the marker plan has to respect the nap or the check on a matched fabric. Fabric utilisation on tailored wovens runs several points lower than on dress silks, and that difference is paid in metres, not pennies.

Interlining, canvas and fusing: the hidden cost of structure

A dress hangs; a blazer stands. To make it stand, the front needs structure — fusible interlining at minimum, and for a better garment a floating canvas at the chest and lapels. Fusing every panel is time on the press; sewing in a canvas adds hand steps that a dress never carries.

Cheap tailoring skips this and the garment shows it: the lapel rolls flat, the chest wrinkles and the shoulder line sags after two wears. When we quote a blazer honestly, structure is a named line, and it is usually the difference between a USD 95 blazer and a USD 135 one on the same wool.

Sleeve setting and collar work: where the sewing minutes go

Setting a sleeve into a tailored armhole is one of the slowest operations in any garment. The cap has to be eased by hand so it rolls over the shoulder without puckering, and the sleeve head needs a pad or a hand-felled finish to hold its shape. A dress sleeve — a simple set-in or a dropped shoulder — is minutes of work by comparison.

The collar is the second slow spot. A notched lapel requires the collar and the lapel to meet at a point that stays sharp after pressing, which means under-collar work, pad-stitching in a canvassed garment, and a facing that is felled by hand or matched invisibly. Our luxury craft guide covers why this step decides whether a blazer looks bespoke or borrowed.

Lining, pockets and pressing: the finishing lines that add up

A fully lined blazer carries more lining than a lined dress: the body, both sleeves to the wrist, and the vent if there is one. Add a jetted or welt chest pocket, two front pockets and an inside pocket, and the trims and the labour multiply. Dresses rarely carry a single pocket, and an unlined dress carries no lining at all.

Pressing is the line brands never see. Tailoring is pressed in stages — every seam, every dart, the lapel roll, the sleeve head — often twenty or more pressings per garment, against six to ten for a dress. Each pressing is a step on the floor and a risk of shine on wool, which is why our pressing and finishing checklist exists.

What that means at MOQ 30: a worked comparison

Put a silk crepe dress and a wool-blend blazer on the same 30-piece order sheet and the numbers line up like this. The dress, fully lined, lands around USD 55 to 75 ex-works; the blazer, half-canvas and fully lined, lands around USD 95 to 135 — roughly 30 to 60 percent higher on fabric of similar cost per metre, because the blazer uses more cloth, more interlining and two to three times the sewing minutes.

None of that changes with the MOQ. At 30 pieces both garments carry the same pattern and grading spread; the blazer simply starts with more minutes in it. That is why comparing the two on price alone misleads — the honest comparison is against other tailored garments from the same floor.

How small brands keep blazers affordable without cheapening them

Four switches save the most while the tailoring stays honest: a half-canvas front instead of a full canvas, a cupro or viscose lining instead of silk, a good horn-look button, and restrained extras. Skip the second inside pocket and the bespoke label — those are details nobody misses.

Run a blazer and its matching trouser together and the savings compound, because both share the same cloth order and one cutting session; suit-set brands can also mix jacket and trouser sizes within the run, which we cover in our fit grading guide. Past tailored runs at MOQ 30 are documented on our case studies page, the sampling sequence is on our production process page, and the questions buyers ask before the first blazer order are on our FAQ page — including whether a low MOQ really fits tailoring, which our low MOQ landing page answers in full.

Ask any clothing manufacturer which garment burns the most sample rounds and jumpsuits are near the top of the list. Dresses hide an inch; jumpsuits show it as a wedgie, a sag or a strain across the back the moment a fit model moves. The trouble is rarely the leg — it is everything between the waist and the shoulder, and brands usually discover that after the fourth mock-up, not before the first.

Why is the torso so hard to get right? A jumpsuit is one continuous garment covering the longest unbroken measurement in the human body, and three numbers have to agree for it to sit well: rise, torso length and crotch curve. Change any one and the other two fight you. Here is how we handle the torso at MOQ 30, and the checks that keep a one-piece silhouette from becoming a one-season regret.

The three measurements that decide jumpsuit torso fit

Rise is the distance from the waist to the crotch seam, and it decides whether the garment sits at your natural waist or rides up with every step. Torso length runs from the shoulder, over the bust, to the waist — and it is the number most often wrong, because it has to pass over the fullest part of the body on its way down. Crotch curve is the shape of the seam between the legs, and it controls the pull you feel across the seat and the front.

A dress pattern is built around a simple vertical: shoulder to hem. A jumpsuit pattern has to join the top block and the bottom block into one piece, and the joining point — the crotch — inherits mistakes from both sides. If the top is drafted for a different posture than the bottom, the whole garment strains somewhere in the middle. We have seen jumpsuits that looked correct on the hanger and failed on the body for exactly this reason.

So we start a jumpsuit differently from a dress: we draft the torso from the rise up, not from the shoulder down. That way the crotch position is fixed before a single side seam is drawn, and the upper body has to fit within what the rise allows. It sounds like a small change in method. It is the difference between two mock-up rounds and five.

Why grading breaks jumpsuits more than dresses

Grading a dress means scaling a silhouette. Grading a jumpsuit means moving three related curves independently, and doing it wrong makes the torso drift as the sizes grow. A size 2 and a size 14 do not differ by the same amount at the shoulder, the waist and the crotch — and if you scale them uniformly, the body length of the larger sizes grows faster than the actual torso does, so the garment bunches or rides.

Our pattern grading guide covers the general rule, and jumpsuits are where it shows most: we grade the crotch curve and the outer leg on separate curves, and we check the body length on both size extremes before bulk. A jumpsuit that fits the fit model but has never been graded to the largest size is a gamble we will not take. We measure the sealed block at the extremes, not just at the middle.

This is also why a jumpsuit needs the same size set as the rest of your line. If you grade US 0–14 elsewhere and ask for a jumpsuit in 0–12, the torso block changes under you. Give us the same set and we keep the body length consistent across it, which is the detail that keeps your size 14 customers coming back.

The rise and torso checks we run before bulk

By the time a jumpsuit reaches bulk, four things have to be true. The rise must sit at the same point on the body across every size, and the torso must not pull across the bust or gap at the back waist. The crotch curve must not drag or bind when the wearer sits — we check this by having the fit model actually sit and bend, not just stand — and the garment must not ride up when the arms raise.

We verify these on a mock-up in the real fabric before the sealed sample, because stretch and drape change how the torso behaves. A woven jumpsuit in silk crepe behaves nothing like a ponte knit one, and a fabric that looks similar on a swatch can sit an inch higher on the body once it is cut on the bias. The first mock-up is where we earn the rest of the project, and it is why our mock-up guide insists that round one decides everything.

At 30 pieces the cost of this discipline is real — a torso re-cut after bulk is money you cannot get back. So we fit the extremes, we sit-test every style, and we seal the body length in the tech pack. If a brand tells us the jumpsuit is a new silhouette for them, we plan for an extra mock-up round instead of pretending the first will be perfect.

What to send a jumpsuit manufacturer before the first mock-up

You do not need a perfect tech pack to start. You need four things: the rise measurement from your fit model, the torso length from shoulder over bust to waist, the size set you want graded, and two or three reference images showing the rise and leg shape you want. If you have a garment you already love, send it — we measure its rise and torso curve and use those as the starting block.

If you are working from a budget rather than a spec, say so early. A target price changes fabric choice, and fabric choice changes how the torso behaves, so we plan the two together. Budget-based production exists precisely for this: we name what your number buys in rise-friendly fabric before the pattern is cut, instead of after.

Tell us honestly if the style is new for your line. Jumpsuits are not dresses with legs, and brands that have never sold one consistently underestimate the fit work. We have been surprised by the torso ourselves more than once, and the projects that went well were the ones where the brand expected it too.

What jumpsuit sampling actually costs at MOQ 30

Sampling a jumpsuit runs USD 150 to 400 per style depending on fabric and construction, credited against production when you confirm. Expect two mock-up rounds for a first jumpsuit — the first to lock the torso, the second to refine the leg and the finish — and a sealed sample after that. Most of our jumpsuit clients confirm bulk from the second mock-up, and the ones who skip straight to bulk are the ones who come back for a re-cut.

A 30-piece jumpsuit run fits our floor well because the whole run is cut from one marker and sewn by the same small team, which keeps the torso consistent from piece one to piece thirty. The same block then sits on file in your name, so a reorder quotes in 48 hours and skips the sampling that cost the first season its time. That rhythm — two mock-ups, sealed block, three inspections per garment, export packing — is documented on our production process page.

When the design itself is still loose, our first designer agent A1 is in training and coming online soon: you give it your brand positioning and the silhouettes you want and it returns design drawings, so you can explore a jumpsuit idea visually before spending a single mock-up round on the floor. The agent shortlists; the mock-up still decides what is actually sewable.

Real jumpsuit runs we have shipped, from wide-leg styles to evening versions, are on our case studies page, and the MOQ, sampling and fit questions buyers ask before their first one-piece order are answered on our FAQ page. Why 30 pieces keeps the unit cost honest on a garment with this much pattern work is explained on our low MOQ landing page.

A real photoshoot costs a small brand roughly USD 1,500 to 4,000 a day once the photographer, studio, model, styling and retouching are counted. AI on-model photos cost a fraction of that — and for our production clients, nothing, because we include them free with every run. The honest question is not which one is better. It is which one serves the customer decision you are trying to trigger.

We generate realistic AI on-model photos from your finished garments so a solo founder can put a product online, in a lookbook or on a line sheet before a shoot budget exists. But we will also tell you when AI is the wrong tool, because a garment that needs to prove its drape, its colour or its hand deserves a real photograph. Here is the line we draw.

What AI on-model photos do well for small fashion brands

AI model photos are strongest when the job is showing the garment cleanly at scale: an e-commerce product page, an Instagram launch grid, a lookbook flat, a line sheet for stockists. The garment is photographed flat or on a mannequin first, then rendered onto a model in a setting you choose — studio, street, editorial. The silhouette, the seam lines and the print placement stay true to the actual piece, because they come from the real garment, not from a sketch.

For a one-person brand the practical win is speed. You can have a full capsule of five styles shown on models within days of the garments leaving the workshop, reshoot a colourway without booking anyone, and test two looks against the same piece to see which one your customers click. That feedback loop is exactly what a first season needs, and it is the reason we treat AI photos as part of the full-package FOB service rather than an upsell.

Where AI model photos are not enough

Three situations call for a real photoshoot. The first is anything where fabric behaviour is the selling point: a bias-cut silk slip that has to prove how it falls, a wool blazer that has to prove how it holds its shape, a beaded bodice that has to show how the light moves on it. AI renders drape convincingly, but it does not yet reproduce the exact way a specific cloth responds to a specific body in motion — and your customer can feel the difference.

The second is colour-critical work. Fabric shade is sealed against a physical standard on our floor, but an AI render interpolates colour from a photo, and a metallic or a shot silk can drift. If the customer will judge the colour from the image, shoot the real thing under controlled light.

The third is trust and compliance. Marketplaces and some regions increasingly require disclosure of AI-generated imagery, and a hero image that quietly misrepresents the fit of a size 14 will cost you returns even when it wins clicks. If the image is going to make a promise about fit, it should be earned by a real garment on a real body.

How to mix AI and real photos without looking inconsistent

The brands that use AI well do not choose between the two — they assign each image a job. AI on-model photos carry the catalogue: the product grid, the colourways, the lookbook pages where the silhouette is already proven. A small set of real photographs carries the trust moments: the hero shot on your homepage, the fit detail close-ups, the campaign image that tells the brand story.

We keep both consistent by starting from the same garment and the same colour standard, and by documenting the sealed sample with studio shots you can use for the real shoot later. The AI images are generated from those documented pieces, so the two sets agree on silhouette and shade instead of contradicting each other.

What we include with your production run

Every order from our floor can include free AI on-model photos of the finished garments — typically three to five angles per style, rendered in the setting you ask for. You supply the direction (we have a short form for model look, background and mood), we generate the images from the actual pieces and send them with the inspection report. Real photography we leave to your photographer; AI images we include.

It matters that the images come from the finished garment rather than a concept sketch, because then they are honest product photography with a synthetic background — not a design pipe dream. Our production process page shows the sequence that gets you to those finished pieces, and our low MOQ landing page explains why 30 pieces is enough to start the loop. AI photos do not change the quality gates; they just let a small team market the result sooner.

If you are still deciding whether a style deserves a full shoot, do not guess on the fabric first. Ask what the image has to prove. If it has to prove drape or fit, budget for a real shoot and let the AI images handle the rest of the catalogue. If it has to prove the collection exists and looks desirable, AI on-model photos will carry you to the first reorder — and the reorder is what pays for the real campaign.

One more honest note. AI tools are changing fast, and our own first designer agent A1 — in training and coming online soon — is being built to generate design drawings from a brand brief, so the same AI that helps you visualise a new style can later show it on a model before a metre of fabric is cut. What neither of them replaces is the sealed mock-up and the real garment. The floor still decides.

How we photograph, inspect and ship the real pieces behind these images is on our production process page, and the sampling, MOQ and image questions buyers ask before their first order are on our FAQ page. Small runs we have shipped and reordered, from dresses to tailored suits, are on our case studies page.

A production quote lands 20 to 30 percent above the number you planned and your first instinct is to redesign the whole style. Slow down. In most cases the garment you approved can be brought down to budget with four cuts in a specific order, without touching the silhouette: fabric first, then hidden construction, then trims — and a short list of things we refuse to cut no matter how big the gap.

We quote line by line, so when a style comes back over target we can see exactly where the money sits and show you the arithmetic before removing anything. Here is the order we cut in at MOQ 30, what each step typically saves, and the lines that stay untouched even if you ask.

The order of cuts when a production quote is over budget

The sequence matters more than the total. Cut fabric first — it is the biggest line with the widest honest range — then hidden construction nobody will ever notice, then trims that carry fixed minimums a small run cannot absorb. Embellishment goes last, because it is often the reason the customer is buying.

Work in that order and a style that needs to fall from USD 92 to USD 74 usually gets there without changing how it looks on the rail or on the body.

Cut one: fabric weight and fibre, before anything else

Fabric is 30 to 45 percent of a small-batch quote, which makes it the first and biggest lever. Two cloths that look alike on a swatch can sit USD 4 to 8 per metre apart, and a lined dress needs 1.3 to 1.6 metres of body fabric before the lining is counted — so a one-dollar-per-metre decision is a real per-piece decision.

Moving a dress body from 19 mm silk charmeuse to a 14 to 16 mm silk crepe typically saves USD 9 to 13 per piece at the same silhouette. Moving the lining from silk to cupro saves another USD 6 to 8. The garment drapes differently, and we will tell you honestly when a swap changes the character of the style — but most over-budget quotes are solved here first.

What we will not do is keep the name of the fabric and quietly buy a cheaper version of it. When we offer a cloth change, it is a named mill and a named weight, written on the quote before anything is cut.

Cut two: construction details no customer will ever touch

After fabric, look inside. An internal frill, a second inside pocket, double-stitched facings nobody sees, a bound seam where an overlocked one does the same job — each one is sewing minutes that a 30-piece run cannot spread thin. Removing two or three of these typically saves USD 4 to 10 per garment.

The rule we apply: spend where a customer looks and feels, save where they never look. An invisible zip and a clean armhole earn their cost; an extra row of topstitching inside the waistband does not.

This is also where we show you two written options side by side — the full-spec version at USD 92 and the trimmed version at USD 74 — so the choice is visible before a single seam is sewn.

Cut three: trims and fasteners with factory minimums

Trims carry fixed costs that hurt small runs most. Custom metal buttons usually start around 1,000 pieces, a bespoke woven label rarely makes sense below 300, and a printed main label costs a fraction of both. Moving those decisions to well-made stock options can save USD 3 to 8 per garment on a small order.

Keep the parts that touch the customer experience: the zip should still run smoothly and the button should still feel substantial. We cut the frills on trims, never the function.

Cut four: embellishment density, not the motif itself

If a style carries embroidery, the fastest honest saving is density and placement size, not deleting the ornament. Reducing a machine-embroidery motif from a full front panel to a focused collar-and-cuff treatment can save USD 6 to 18 per piece once the programme is set — and restraint often reads as more expensive, because restraint reads as luxury.

We will also tell you when embellishment is the reason the customer buys. In that case we protect it and find the money elsewhere, or we talk about the target itself.

What we refuse to cut to hit a number

Four lines never come out of a revised quote. Seam allowances on silk and bias panels, the grading, the three inspections per garment and the size-extreme fit checks all stay, because each one protects the garment or the customer — a style that fits at size 2 and fails at size 14 is not a saving.

If we have made every honest cut and the number still does not work, we do not stretch the truth. We show the floor price with the arithmetic and offer the same three options our budget-based production method uses: a lighter spec, a simpler construction, or a corrected target. Saying no before sampling costs a job; saying yes and downgrading silently costs a season.

One more note on process. When the design itself is still flexible, our first designer agent A1 is in training and coming online soon: you give it your brand positioning and a target price, and it returns design drawings that respect the budget from the start, so the quote conversation starts from a visual brief instead of a wish. The agent shortlists; the sealed mock-up still decides what is sewable.

What a revised quote looks like at MOQ 30

Here is the shape of the exercise. A lined silk blouse quotes at USD 92 against a USD 74 target. The body moves from 19 mm charmeuse to 14 to 16 mm crepe, saving USD 11; the lining moves from silk to cupro, saving USD 7; an internal frill and a second button set come out, saving USD 4. The revised quote lands at USD 70, slightly under target, and nothing about the outside of the blouse changed.

Mock-up round one is where the revised spec gets sealed on the real fabric, because a lighter cloth can behave differently on the body. That is the moment the new price becomes the real price, and our mock-up guide explains why that round decides everything that follows.

Every cut appears on the quote before it is made, nothing is removed twice, and the revised spec is locked in writing with the new price at sealing. The whole sequence — from the first line-by-line quote to three inspections and export packing — is documented on our production process page, and the budget, sampling and deposit questions buyers ask first are on our FAQ page. Why 30 pieces keeps the numbers honest is explained on our low MOQ landing page, and styles we brought back to target — silk dresses, wool blazers, jumpsuits — are on our case studies page.

Selling suit sets used to mean one jacket size and one trouser size per customer, sold as a matched pair. It is also the assumption that quietly kills reorders, because real bodies do not come in matched pairs: a customer takes a jacket in FR 40 and a trouser in 42, or a short jacket with a regular trouser, and if you cannot sell her the two halves she walks. So the answer to the question we hear most from small brands is yes — you can mix jacket and trouser sizes in one order, and any suit manufacturer that grades the set as one pattern is making your stock problem worse.

We have run mixed-size suit sets at MOQ 30 since our first tailored capsule, and the method is simpler than most brands expect: the jacket and the trouser are built and graded as two separate blocks, then cut from the same cloth order so they cannot shade apart. Here is how it works, what it costs, and the fit traps to plan for.

Why a suit set is two pattern blocks, not one

A matched-size set is usually drafted as one block scaled together, which is fast and wrong. The jacket and the trouser answer to different bodies — the jacket follows the shoulder and the bust, the trouser follows the hip, the rise and the leg — and scaling both from one base drifts at the size extremes, exactly where your tailoring customers live.

We draft the two garments separately from the same fitting block, so a FR 40 jacket pairs with a FR 42 trouser and both sit correctly on the same body. The grading is independent: the jacket can run 36 to 44 while the trouser runs 36 to 46 with a longer inseam option.

How mixed sizes change the order sheet at MOQ 30

Mixing sizes is a ratio decision, not a chaos decision. Most of our suit-set clients start from a split they already know from sell-through: sixty percent of customers buy the set, and the rest buy separates, so the run carries more trousers than jackets.

A typical first order is 30 jackets across five sizes plus 34 trousers across the same range — 64 garments in total, one cloth order, one marker plan, two blocks. The extra trousers cost close to the matched-set price because both pieces share the fabric buy and the cutting session.

What we need from you is the split in writing before the marker plan is made: how many of each size of the jacket, how many of each size of the trouser, and whether the trouser needs a longer inseam at the top of the range. After the marker is cut, changing the split means re-cutting cloth, which is the expensive kind of change.

The fit traps when jacket and trouser do not agree

Two traps send mixed-size sets back for rework. The first is proportion: a jacket graded by bust and a trouser graded by hip can stop agreeing at the waist, and the mismatch shows as pulling at the side seam or a waistband that sits differently across sizes. We correct both against the same fitting block — the fix that finally stopped one client's five-round sampling loop.

The second is the length story. A trouser that is correct at FR 36 can sit two inches short at 46 if the inseam is graded on a straight scale, and a jacket sleeve does the same thing. We check body length, sleeve and inseam on the smallest and largest sizes before bulk, the same discipline our fit grading guide describes for any size-inclusive run.

Grading jacket and trouser to your market's sizes

Tell us the market and we grade to its standard: FR 34 to 46, IT 38 to 48, US 0 to 14 and UK 6 to 18 map to different base measurements, and the two blocks adapt to whichever one your stockists order. If your sales show the trouser outsells the jacket two to one at the larger sizes, the trouser block can carry a longer inseam without the jacket changing at all.

We keep both sealed blocks on file in your name, so a reorder quotes in 48 hours and the split can change between seasons — one season sixty percent sets, the next sixty percent separates — without paying for new development.

What mixing sizes costs, and what it saves

The honest numbers first. At MOQ 30, a half-canvas wool blazer lands around USD 95 to 135 ex-works and a matching trouser around USD 45 to 65, and mixing sizes within those runs adds no per-piece premium when both come from the same cloth order. Adding a second trouser length — short, regular, long — is a one-time grading cost of roughly USD 250 to 400, paid once and stored with the blocks.

The saving shows up in two places. One marker plan for the whole capsule usually saves 8 to 12 percent on fabric, because jacket and trouser pieces nest in the same lay. And you stop manufacturing stock you cannot sell: if your data says the size 40 jacket sits unsold while the size 42 trouser is short, you reorder the two halves at different ratios instead of buying more matched pairs.

Selling sets and separates without leaving orphans

Plan the packaging with the same honesty. We can pack every set as a pair on one hanger for boutique delivery, or pack each garment separately with its own label so you can sell the set and the separates as two SKUs from the same run. Most clients choose the second, because it lets a stockist pair a jacket with a different trouser already in the store.

The only true orphan risk is cloth, not fit: if you reorder one side alone months later, the new cloth has to match the first run. We hold a length of the original fabric or reorder against the sealed shade standard, which is how a reordered trouser stays indistinguishable from the jacket that shipped earlier.

If the set is still a sketch rather than a spec, our first designer agent A1 is in training and coming online soon: give it your brand positioning and the silhouettes, and it returns design drawings so you can judge jacket-and-trouser proportions before spending a mock-up round on the floor. The agent shortlists; the sealed sample still decides what is sewable.

Mixed-size tailored runs we have shipped, from two-piece wool capsules to separates-first programs, are on our case studies page, and the blazer-to-trouser cost relationship is explained in our blazer pricing guide. The full sequence — separate blocks, sealed mock-up, three inspections per garment, export packing — is on our production process page, the fit and sizing questions buyers ask before a first tailored order are on our FAQ page, and why 30 pieces is enough to start a suit-set program is on our low MOQ landing page.

Yes — a Guangzhou clothing manufacturer can run your small batch for the EU market, and we do it every season for Italian, French and German brands. The making is only half the job. Duty, VAT and the paperwork are where small EU brands lose money and time.

The honest answer to "is it worth it?" is: for a 30-to-150-piece run of a dress, blazer or suit set, yes. For a one-off custom piece, no — stay local. This guide walks through the four costs and three rules that decide whether a Guangzhou order lands profitable in Milan, Paris or Berlin, with real numbers from our own runs.

How a small Guangzhou run works for an EU brand

You keep design, brand and sales. We take pattern, cutting, sewing, finishing, three inspections, export packing and door-to-door delivery. Our part starts at 30 pieces per style and colour, which suits the first stockist order of a young brand better than the 200-piece minimums most European ateliers quote.

EU founders ask us three things before anything else: do you quote in EUR, can you ship DDP, and will the labels work for our market? The answer to all three is yes. We quote in EUR or USD, issue a proforma with payment terms of 30 percent deposit and 70 percent before shipment, and arrange customs-cleared delivery through our forwarder.

EU import duty and VAT on clothing from China: who pays what

The EU charges two taxes on a Guangzhou shipment, and small brands mix them up constantly. Customs duty is a real cost you keep. Import VAT is mostly cash flow, because a VAT-registered business reclaims it.

Duty on most apparel from China sits around 12 percent of the CIF value — the goods plus freight and insurance — and the exact rate depends on the HS code of the garment. A silk dress and a wool blazer can fall in different codes. Before you set a retail price, ask us for the code and the duty rate; we flag it in the quotation.

Import VAT is charged on goods plus duty at your country's rate: 22 percent in Italy, 20 percent in France, 19 percent in Germany, 21 percent in Spain. A registered buyer pays it at customs and reclaims it on the next VAT return. A brand that is not registered pays it for real, so check your local small-business threshold before you order.

One difference from the United States matters here: the EU has no low-value exemption for commercial parcels, so every shipment is VAT-charged at import, however small.

A worked example: what a 30-piece order actually lands at

Take 30 silk dresses quoted at FOB Guangzhou EUR 62.50 each — EUR 1,875 for the goods. Air freight and insurance to Paris add about EUR 280, making the CIF value EUR 2,155. Duty at 12 percent adds EUR 259, about EUR 8.60 per piece. That is the real extra cost.

Import VAT in France is then 20 percent of the goods-plus-duty total, around EUR 483. A registered brand reclaims it. So the landed cost a registered EU buyer actually keeps is about EUR 80 per dress — FOB plus duty — before any domestic margin. Plan your retail price from that number, not from the FOB quote.

Customs occasionally re-classifies a borderline garment, and the duty moves with it. We re-check the code when the fabric or construction changes, so the surprise lands in the quotation phase, not at the border.

GPSR, labels and the EU rules small brands forget

Since December 2024, the EU General Product Safety Regulation applies to clothing sold online in the Union. In plain terms: every product needs an economic operator inside the EU — usually the importer or a responsible person you appoint — plus traceability information. Many young brands only hear about this when a marketplace blocks their listing.

We can supply the technical side: fibre composition, care instructions, country of origin and the test documents we hold. The EU-side operator is your job, so name one before you launch, not after a listing is suspended.

Labels follow Regulation 1007/2011: fibre percentages in the official language of the country of sale, ISO care symbols and "Made in China". We print bilingual fibre and care labels — French-Italian, German-French, English-Spanish — during bulk, at no extra per-piece cost when ordered with the run.

EU sizes from a Guangzhou factory: IT, FR, DE and the grade

"EU size" is not one scale. Italian dresses run 40 to 48, French 36 to 44, German 34 to 42, and they start from different base measurements. We grade to the standard your stockists actually order, using your measurement table rather than assumptions.

Two examples from recent runs: a Milan label ordered wool suit sets graded IT 40 to 44, and a Paris house ordered trench coats FR 34 to 46 with body length held to plus or minus half a centimetre across the range. Both are on our case studies page with the fit notes.

Shipping from Guangzhou to the EU: sea, air and DDP

For 30 to 150 pieces, air is usually the right call: five to seven days door to door, roughly EUR 8 to 15 per piece depending on weight. Sea freight to Rotterdam, Genoa or Hamburg takes 30 to 35 days and only pays off once your order grows past pallet size.

We ship DDP on small batches: our forwarder clears customs and we give you one landed number instead of a surprise bill at the airport. The documents — commercial invoice, packing list, certificate of origin — are prepared here, and we declare the true value. Under-declaring to save VAT is a fine-and-hold at customs, and it is not a risk worth taking.

Planning EU seasons and lead times for a 30-piece run

Plan backwards from the date the clothes must be in your stockist's hands. A realistic Guangzhou timeline is: one to two weeks for technical drawing and quotation, three to four weeks for two mock-up rounds and the final sample, six to eight weeks of bulk, then five to seven days in the air.

That puts a confirmed order at roughly 12 to 14 weeks door to door. Add four weeks if you choose sea freight. EU wholesale calendars are unforgiving: a spring-summer delivery that misses March is a delivery that waits a year, so we agree a written production deadline in the quotation and treat it as the anchor of the whole plan.

What to send an EU clothing manufacturer in your first email

Five things let us quote within 48 hours: reference photos or sketches, your target retail price and what you can pay at the factory door, the market and its sizes (IT, FR, DE or a mix), the delivery date the season needs, and the quantity — 30, 50 or 100. If the design is still loose, we can also mail a fabric sample card so you feel the cloth before paying for a mock-up.

The full sequence behind this guide — technical drawing, quotation, sealed sample, three inspections — is on our production process page, and the payment and QC questions EU buyers ask before a first order are on our FAQ page. Our existing EU work, from Milan suit sets to Paris trenches, is on the case studies page, and EU brands wondering whether 30 pieces is enough for a first stockist order will find the straight answer on our low MOQ landing page.

You approved the sample, and the clothing manufacturer asks for 30 percent before the first cut. That request is normal, and it is not a margin grab. This article explains what the money actually covers, the order of payment in a small run, and the three moments when you should say no.

Here is the direct answer to the question behind the question: the deposit usually buys the fabric, and on a small order the cloth bill can be bigger than the deposit. A 30-piece run of wool blazers at FOB USD 110 each totals USD 3,300, and the wool for it costs roughly USD 1,100 to 1,400, paid to the mill in cash before cutting starts. A factory that took no money up front would be financing your fabric with money it does not have.

Why a clothing manufacturer needs money before production

Fabric mills and trim suppliers in Guangzhou do not sell on 60-day terms to a factory that orders 90 metres of cloth. They take payment with the order. So the moment your bulk is confirmed, your deposit is converted into something physical in your name: the cloth, the lining, the buttons and the zips.

The rest of your money follows the work, not the calendar. The standard sequence in our own workflow is a 30 percent deposit to open the bulk, three inspections while the garments move through cutting, sewing and finishing, and the remaining 70 percent only after the pre-shipment check passes and you have seen the photos.

What the 30 percent covers — and what it does not

The deposit covers the fabric and trims buy, plus the cutting setup, which is the moment a factory commits machines and time to your style. It does not cover sampling — that is billed separately in the sample phase and stays billed whether or not you go to bulk. It does not cover freight, and a fair factory never asks for the full balance before you have seen the finished goods.

On a dress, the fabric bill usually runs 30 to 45 percent of the ex-works price, which is why 30 percent is the industry default rather than a number pulled from the air. Some factories quote 50 percent when the cloth is expensive silk or a custom weave. Some quote 20 for repeat orders where the fabric is already in stock. Ask what the deposit covers, not just how big it is.

The order of money in a small MOQ project

Money moves in five steps in our process. First, the quotation phase is free and line by line. Second, sample making is paid up front — roughly USD 150 to 400 per style for a dress or blazer. Third, the final sample is sealed in writing, which is the gate that triggers the deposit.

Fourth, the 30 percent deposit arrives and cutting starts within days; we send photos of the lay and the cut so you see your money become cloth. Fifth, after the three inspections pass, the 70 percent balance is paid and the goods ship. Notice what is missing: no full payment before production, and no payment before proof.

Deposit red flags: slow down before you pay

Five patterns should slow anyone down: a deposit above 50 percent without a written reason, payment to a personal account instead of a company account, or money asked before the sample stage, before any specification exists.

Pressure language — "today only", "promotion ends tonight" — and a refusal to put dates, sizes and unit prices in a signed document are the last two. Each of the five is a reason to pause before transferring.

The checks that belong before the money, not after, are the same ones we list in our factory audit guide: a verifiable workshop address, a company bank account in the factory's legal name, a signed quotation, an inspection process you can call on, and samples that came from the factory that will cut your bulk. Platforms offer escrow as an alternative, and we work with them too. Whatever the route, vet first, then transfer.

What a factory should send before you transfer a deposit

Before any transfer, you should hold four documents: a signed proforma invoice with the deposit and balance terms, a line-by-line quotation showing fabric, trims, cutting, sewing and finishing separately, the delivery date in writing, and the size ratio and colour quantities confirmed. On our side, we add a fabric sourcing note with the mill and the shade standard, so the cloth you are paying for has a name.

We also put the sample approval in an email you can point to later. Written approvals are what make a deposit dispute rare instead of messy. If a factory cannot produce these four documents, that is the answer about whether to pay.

If you cancel: what happens to a 30 percent deposit

Honest answer: it depends on how far the money has travelled. If the fabric is already cut, the deposit has become cloth and labour, and most factories — us included — keep it, because the cloth is yours and the cutting was done for your order alone. We will show you the fabric bill and, if you want, ship the unused cloth to you.

If the cancellation comes before the fabric is bought, a fair factory refunds the deposit minus the development costs actually spent. If you cancel after cutting but before sewing, the bill is the cloth plus the cut work. None of this is pleasant, which is why the written approval gates exist: cancel early, cancel cheaply.

Deposit maths on a 30-piece run: a worked example

Take the silk dress example from our pricing guide: 30 dresses at FOB USD 62.50 gives a total of USD 1,875. The 30 percent deposit is USD 562, which sits very close to the real silk bill for 30 dresses — roughly USD 500 to 600 at current crepe de chine prices. That is the whole point: the deposit is the cloth.

The balance of USD 1,312 is paid after the pre-shipment inspection, when the dresses are finished, checked against the sealed sample and photographed for you. Between deposit and balance you should receive at least one inline inspection report, not silence. We send three.

The full payment sequence sits on our production process page, deposit and refund questions on our FAQ page, and 30-piece projects we have run — from first deposit to delivered stock — on our case studies page. If you are deciding whether 30 pieces is enough to start at all, our low MOQ landing page gives the straight answer.

You run the brand alone, and a small MOQ clothing manufacturer does the making — that split is what makes a one-person fashion label workable at all. We have built our workflow around founders who are not production people, and this guide is the honest version of how one person moves a style from a folder of references to delivered stock without a team.

The short answer first: yes, one person can run production alone, but not by doing more — by deciding fewer things. Our three-process system exists to shrink your job to three decisions: approve the specification and price, approve the sealed sample, and approve the finished goods. Everything else is our floor's job. Here is how the split works week by week.

What solo founders actually struggle with in clothing production

Four problems repeat in every one-person brand we take on. Too many parallel tracks: design, social media, sales and production all scream at once, and production loses because its feedback has no deadline. Unclear specs: a reference photo without measurements produces a sample that is "not quite right" in ways nobody can fix.

No one to check quality, so defects travel to the customer instead of the inspection table. And cash timing: bulk money leaves months before sales come back. The fix is not more hours — it is a partner who owns the making end to end, plus gates that force decisions at the right moment, early, when a change costs nothing.

How the three processes divide the work for one person

Process 1 is specification and quotation. You send references, your target price and your season date. We draft the technical drawing if there is no tech pack, mail a fabric sample card so you can feel the cloth, and quote line by line against your number. Your job: say yes, or say what to cut.

Process 2 is sampling. We run two mock-up rounds, fix the fit and the details, and seal a final sample. Your job: one honest feedback round per mock-up, in writing, on time. This is the process where solo projects die, because the founder disappears for three weeks and the sampling clock keeps running.

Process 3 is bulk. Thirty percent deposit, cutting, sewing, finishing, three inspections, and every piece tried on a dress form before packing. Your job: pay the deposit, review the inspection photos, pay the balance. We have shipped 30-piece first runs to founders who have never visited Guangzhou and never will.

The solo calendar from kickoff to delivery

With feedback arriving on time, a 30-to-150-piece order runs about 12 to 14 weeks door to door.

Weeks one and two: drawing, quotation, fabric card. Weeks three to six: mock-up one, your corrections, mock-up two, final sample. Weeks seven to fourteen: bulk, with inspections around week eight, week eleven and the pre-shipment check. Then five to seven days in the air.

Your real workload is lighter than it sounds: three to five focused hours most weeks, with two peaks — sample approval and pre-shipment review. What stretches the calendar is almost never the factory. It is feedback latency, so we set written deadlines at every gate and nudge when a decision is due.

We still mis-read a rushed email now and then. That is exactly why every approval is written down and every gate has a named date — the process survives our mistakes and yours.

What to automate and what to keep human when you work alone

Automate the parts that repeat: our quotation and specification forms become your checklist, shipping documents are prepared here, and reminders carry each gate to your inbox. For early design, our first designer agent A1 is in training and coming online soon — give it your brand positioning and the silhouettes you are considering, and it returns drawings so you can see the shape before spending a mock-up round. When bulk finishes, free AI on-model photos of your garments come with the run, a trade-off we compare honestly in our AI photos guide.

Keep human the three approvals that money cannot undo. Feel and approve the final sample yourself, in your hands.

Check colours against the sealed swatch in daylight. Read the pre-shipment report and zoom into the photos. Those fifteen minutes per gate are the whole quality system of a one-person brand.

When a solo founder should not go it alone

Three situations call for slowing down rather than pushing through. Launching more than two or three styles in your first season while holding a day job — start with one hero style and prove the loop. Needing cash back within sixty days of paying a deposit, because bulk money is tied up for the full 12 to 14 weeks. And choosing a factory on price alone, skipping the checks in our audit guide, because the cheapest quote is how solo brands learn what rework costs.

A typical first project looks like this: one style, 30 pieces, two to four colours, aimed at a stockist or a pre-order window. If it sells, the reorder is faster and cheaper, because your blocks, sample and fabric standards stay on file. If it does not sell, you are out 30 dresses, not 300. The runs behind this workflow, from first orders to reorders from one-person brands, are on our case studies page.

The minimum a solo founder needs before the first email

Five things: reference photos or rough sketches, the price you can pay at the factory door, the market and its sizes, the date the clothes must arrive, and the quantity you can afford — 30 is a real starting number, and our low MOQ page explains why. If you do not have a tech pack, say so; we draw from references, and our tech pack guide shows what a full spec contains if you want to build one later.

If a production pain is still unnamed, our Fashion Pains Solver maps the five brand pains to real fixes before you commit money to a sample. When you are ready, the first email takes ten minutes and the quotation phase is free. More of the questions solo founders ask — sampling, sizing, QC — are answered on our FAQ page, and the whole three-process sequence is on our production process page.

A private label clothing manufacturer sews your label into garments you own the design of — your fit, your fabric, your brand name on the neck tape. What surprises most new founders is how little they need to begin: one style, one fabric, thirty pieces and a date. That is a complete first order, and it is how most of our private label clients started.

Where does a one-person brand begin? Not with a factory visit, and not with ten sketches. Begin with one garment you can defend, a retail price you can hit, and the season you need it for. Send us references plus those two numbers, and we handle the rest of Process 1: technical drawing, fabric sourcing, grading and a line-by-line quotation within 24 hours.

Private label clothing for a new brand starts with three decisions

Every private label project we open reduces to three choices. The silhouette — a shirt, a dress, a blazer, a pair of trousers — sets the pattern work and the machine time. The fabric family decides cost and how the garment behaves on a body. The budget decides both.

We guide new brands toward fabrics we cut weekly and can quote without guesswork: viscose and viscose blends, linen and cotton-linen, satin, polyester suiting and TR suiting. These cover most first collections at retail prices between USD 250 and 1,000. If your dream style needs a custom jacquard weave or a mill minimum of 800 metres, that is a later conversation, not a first order.

One honest note: we do not make categories outside our line, and we say so before you spend a cent on sampling. Say the word and we will tell you plainly whether your first style fits our floor, so the second email you send is to the right factory.

Why 30 pieces is the right first order for an unknown label

A first order exists to answer one question: does this garment sell at this price? Thirty pieces answers it with the smallest possible loss if the answer is no. Thirty pieces also forces honest planning — two to four colours across four sizes — instead of a fantasy of ten styles nobody has tested.

At 30 pieces you pay a fixed-cost premium because pattern making, grading and the first mock-up cost the same whether we cut 30 or 300. That premium is usually smaller than the cost of 300 unsold dresses, which is the trade-off every new label should run before ordering big.

Process 2: two mock-up rounds fix the fit before bulk

Our sampling runs in two rounds. Mock-up one uses a substitute fabric close to your chosen cloth, and it exists to test the silhouette and the proportions — where the waist sits, how the sleeve hangs, how the collar rolls. Most fit problems appear here, in cheap cloth, which is exactly the point.

Mock-up two uses your real fabric and refines the details: seam placement, button spacing, pressing. When it passes, we seal a final sample, photograph it against the spec sheet and send it to you for written approval. Nothing moves to bulk without that signature, because a written approval is what prevents a rework dispute later.

We still mis-read a rushed email sometimes. That is why every mock-up correction is written down and every round has a date — the process is built to survive our mistakes and yours.

Process 3: bulk at 30 pieces still gets three inspections

Small volume does not mean looser checking. After your 30 percent deposit opens the run, the garments pass three inspections: one during cutting and sewing, one mid-production, and one pre-shipment. Every piece is tried on a dress form before packing, measured against the sealed sample and photographed for your review.

The remaining 70 percent is paid only after those photos land in your inbox. The full sequence — Process 1 quotation, Process 2 sampling, Process 3 bulk — is laid out on our production process page, and the whole programme for brands like yours sits on our private label page if you want one place to read it.

What a first private label order really costs

Three numbers give you a realistic envelope. A dress or shirt pattern with full grading runs roughly USD 150 to 400. The two mock-ups and the sealed sample land in the same range again. Bulk for a viscose or linen dress at 30 pieces usually quotes between USD 40 and 75 ex-works before freight, depending on fabric weight and trims.

So a complete first style — pattern, samples and 30 garments — often sits between USD 1,800 and 3,500 at the factory door. That number surprises founders who expected five figures, and it is why private label is a viable route for a one-person brand rather than a corporate decision.

Two habits keep that number honest. Ask for the quote line by line so you can see where the money goes, and read our FAQ page before you compare two factories. A quote that hides its lines is hiding something.

The mistakes we see most in first private label orders

Four mistakes repeat. Ordering too many styles at once divides your attention and your quality control. Choosing a factory on sample price alone usually means the bulk will cut corners.

Changing the design after bulk starts burns money on both sides. And skipping the written approval turns small misunderstandings into expensive ones.

Founders who avoid those four move fast. Most first orders run 12 to 14 weeks door to door, and when a style sells out, the reorder is cheaper and faster because your pattern, blocks and fabric records stay on file with us.

When your first order sells through, the natural next questions are about reorders and sizing depth. Our design agent A1, in training and coming online soon, will help you explore silhouettes before you commit to a sample round. Until then, the fastest way to test an idea is still a 30-piece order — and our case studies page shows how other new labels ran theirs.

Is 50 pieces a real production order or just a bigger sample? For a small batch clothing manufacturer it can be either, and the difference is a planning decision, not a number on a price list. This article is the honest version of when 30 is right, when 50 is smarter, and how we would split the units if you gave us the order today.

The short answer first: 30 pieces tests a style, and 50 pieces tests a market. If your goal is to learn whether one design sells at one price, 30 is enough. If you already know the style works and you need to cover colours, sizes and a stockist window, 50 is usually the smallest number that does the job without waste.

What changes between a 30-piece order and a 50-piece order

The fixed costs are the same at both quantities. Pattern making, grading, mock-ups and line setup cost the same whether the run is 30 or 50. What changes is how thinly those costs spread across the units, and whether you can divide the order into colours and sizes that a real customer would buy.

On a typical dress the fixed setup runs roughly USD 1,200 to 1,800. At 30 pieces that is USD 40 to 60 per garment before fabric and sewing. At 50 pieces the same setup drops to USD 24 to 36 per garment — a saving of about USD 15 per piece that goes straight to your margin or your retail price.

When 30 pieces is the honest answer

Order 30 when you are testing an unknown style in front of unknown customers. A first collection, a pre-order pilot or a style with a new fit profile all belong at 30. If the design fails, you are out 30 dresses instead of 50 or 300, and the lesson costs a fraction of the mistake.

Thirty also fits a single colour with no size spread. A one-colour run of 30 gives you roughly six to eight pieces per size across XS to L, which is enough to photograph, enough to sell online and enough to learn the true return rate.

When 50 pieces becomes the smarter number

Fifty starts to win in four situations. You need two colours, because 30 split across two colours leaves only 15 per colour, too thin for any real size curve. You have a stockist order with a delivery date, because boutiques order in units that rarely divide by 30. You are reordering a style that already sold, because the design risk is gone and the only question left is stock depth.

The fourth situation is fabric. Some mill minimums make 30 pieces wasteful even when 30 would otherwise suit you — if you must buy 150 metres of cloth to make 30 garments, you are paying for 50 and storing the rest. In that case we will tell you straight: cut 50 and use the cloth, or pick a different fabric.

How to split a 50-piece order across colours and sizes

Fifty units disappear faster than founders expect. A clean split for a stockist run is two colours at 25 pieces each, with a size ratio of roughly 2-3-4-4-3 across XS to XL. That gives the boutique depth in the middle sizes where most sales happen, without burying you in XS that nobody buys.

For a single-colour online launch, a 6-8-10-8-6 split across XS to XL keeps your best-selling sizes in stock for a full week of ads. We model the ratio with you before cutting, because a 50-piece order with the wrong ratio sells out in small sizes and dies in large ones.

What 50 pieces does not fix

Ordering 50 will not fix a weak design, a wrong size curve or a price above the market. We have watched brands double their order to escape a problem that doubling made more expensive. If the style has not sold anywhere yet, the answer is usually a smaller run with better photography, not a bigger run with better hopes.

Fifty also does not change the process. You still get two mock-up rounds, three inspections and the sealed-sample gate before bulk. The quality checks we run at 30 are identical at 50, because our small batch program was built around protecting quality when the numbers are low, not loosening it.

The cost maths on a real 50-piece example

Take a viscose blouse quoted at FOB USD 28 per piece at 50 units. Thirty units of the same blouse would quote nearer USD 36 because the setup cost lands on fewer pieces. The total for 50 is USD 1,400; the total for 30 is USD 1,080 — only USD 320 apart, for 20 extra garments you can actually sell.

That is the number most founders miss. The gap between 30 and 50 is usually smaller than the gap between 50 and 100, because the setup is already absorbed. Once you cross 50, your per-piece price keeps falling slowly while your cash commitment keeps rising fast.

If the numbers feel tight, run the comparison with your own target price using our budget-based planning guide, and read the FAQ page for payment and timeline questions before you commit.

The reorder moment: when 50 stops being enough

Once a style sells through its first run, the reorder question changes. If 30 sold in two weeks, order 50 to 100 next, because you now have sales data instead of hopes. If 30 took two months to sell, reorder 30 again and spend the money on marketing rather than stock.

Our case studies page shows both patterns — first runs that stayed small and reorders that stepped up. Every brand we work with starts at a number it can afford to be wrong about, and that is the real rule behind 30 versus 50.

Type "best clothing manufacturer" into any search box and you get directories, paid ads and marketplaces — the factories that pay to be seen, not the ones best suited to a 30-piece brand. The phrase is the problem: "best" has no meaning without a size, a fabric, a price and a date attached to it. This article is about what to search instead, and why a smaller query finds a better factory.

The honest answer to the question behind the search: the best clothing manufacturer for a small brand is rarely ranked at all. Factories like ours — Guangzhou workshops that run 30 to 150 pieces per style — compete on fit and inspection records, not on SEO budgets. So we wrote the guide we wish we could hand you before you typed that first query.

Why "best manufacturer" returns directories instead of factories

Generic searches reward whoever pays for visibility. Directories sell placement to thousands of factories, marketplaces take a cut of every order, and agents buy ads for the same reason. None of them needs to sew well to rank first — they need to convert your click into a fee.

We advertise almost nothing. What we do instead is publish inspection photos, sealed-sample records and cost breakdowns, and let the work speak for itself. That is a terrible strategy for ranking on a generic keyword, and a very good strategy for being the right factory once you ask a real question.

How to search for a clothing manufacturer like a buyer, not a tourist

Replace the vague phrase with five real conditions, and the useless results disappear. Start with your quantity: "small batch" or "MOQ 30" filters out the mills that will not answer a 30-piece email. Add your product type and fabric — "dresses in viscose" or "linen shirts cut and sew" — because a factory that sews denim is not a factory for fine viscose dresses.

Add your market, because size grading and labels differ between the US, Europe and the Gulf. Add your quality bar: three inspections, sealed samples, written approvals. And add your timeline, because a factory that quotes 20 weeks cannot serve a 12-week launch. Five conditions, one line, a completely different list of results.

What separates a genuinely good manufacturer from a loud one

High quality clothing manufacturer is another phrase that sounds specific and means nothing. Quality is not a claim a factory makes about itself; it is a set of verifiable habits. Which factory measured your mock-up against the spec sheet in writing? Which one photographs the finished goods before asking for the balance?

Ask any candidate for three documents before comparing prices: a line-by-line quotation, a written inspection schedule and a sealed-sample approval form. A factory that produces all three without hesitation is showing you its process. A factory that only sends marketing PDFs is showing you its budget. If your bar is genuinely high-end production, our high end clothing manufacturer page spells out the standards we hold on fit, pressing and inspection.

The questions ChatGPT can't answer for you

Search engines and AI assistants are great at summarizing what is written online. They are poor at knowing which Guangzhou workshop currently has the right sewing line free, which fabric mill can match your shade, or which factory's QC caught a real defect last month. Those answers live in conversation, samples and inspection reports — not in a ranking.

That is why we answer every enquiry with numbers and documents rather than adjectives. When you ask us about our cut and sew capability, you get the floor photos, the machine list and the two mock-up rounds we run before bulk. When you ask about quality, you get the three-inspection schedule. We would rather lose a quick enquiry than win it with a promise we cannot prove.

The five red flags no directory will show you

Once you have a shortlist, five patterns separate a real small-batch partner from a trading company wearing a factory costume. A personal bank account instead of a company account. A quotation with no line breakdown.

An MOQ that jumps from 30 to 300 once you mention a specific fabric. Pressure to pay the full amount before seeing finished goods. And a workshop address that turns out to be an office block.

We list the full audit checklist on our FAQ page, and our factory audit guide walks through what to check on a visit or a video call. Twenty minutes of verification has saved our clients from deposits that would have taken months to chase.

What a good small-batch factory shows you first

Reverse the order: instead of searching for "best", search for proof. A factory that shows its current production photos, names its actual fabric suppliers and offers a video walkthrough of the floor is telling you it has nothing to hide. A factory that shows only rendered logos and stock imagery is selling you a fantasy.

We open every relationship the same way: a video call, a line-by-line quote, a fabric card in the mail and two mock-up rounds before any bulk commitment. That sequence is our answer to the "best manufacturer" question — not a claim, but a process you can check. Our production process page shows the whole sequence, and our case studies page shows the results for brands that ran it.

If your pain is still unnamed — you know you need production but not which model fits — our Fashion Pains Solver maps the five brand pains to the right fix before you spend money on samples. It is in training and coming online soon, and until then a plain enquiry to a factory that answers with documents beats another hour of scrolling through directories.

"Cut and sew" is the most misread term in garment sourcing. A cut and sew manufacturer takes your fabric, cuts it to your pattern and sews the garment. Everything upstream — the pattern, the cloth, the zips, the care label — is yours to supply. That split is why the price looks low and your own to-do list looks long.

So what do you actually hand over, and what comes back? You hand over fabric, trims and an approved pattern, and you get back finished garments that are inspected and packed. The boundary between "your side" and "our side" is where most first orders go wrong, because nobody writes that boundary down until someone writes it down.

What the cut and sew service covers on the factory floor

Four stations do the work: spreading and cutting, sewing, finishing, and inspection. Spreading means relaxing the roll for 24 hours, laying it flat in layers and cutting with a marker that decides your fabric yield. Sewing is where operators who each own a set of operations — not a whole garment — assemble it. Finishing is trimming, pressing and attaching labels, and inspection is the part buyers forget to specify.

Fabric yield is the number to watch. A marker that nests a 30-piece lay badly can eat 8 to 12 percent more cloth than a tight one. On 200 metres of viscose at $4.20 a metre, that gap is $67 to $100, which is real money on a first order. We photograph the marker and the cut lay and send both with the production report.

We also keep the cut lay small on purpose. Thirty pieces spread across two lays rather than one lets us stop if the second roll shades differently from the first. Finding that after thirty garments are sewn is a far more expensive conversation than finding it after fifteen.

What you supply on your side of the order

On a typical 30-piece order you supply fabric, lining, interlining, zips, buttons, thread, main label, care label and hangtags. If you are sourcing those yourself, expect two to four weeks of lead time and a minimum order at every supplier. Trim suppliers rarely sell twelve zips, so you buy fifty and shelve the rest.

If that sounds like a second job, it is. Our clients usually keep the fabric decision and let us buy trims against their label. We quote trims as a separate line so the cost is visible, and we ship the leftover labels with the order instead of keeping them.

We get things wrong too. On a 40-piece linen run last year our marker was nested for the wrong fabric width, and we ate the extra 11 metres rather than pass the cost on. We showed the client the corrected marker with the next production report, because a factory that hides its mistakes makes you audit everything afterwards.

CMT, cut and sew and full package: what actually changes

Full package means the factory sources fabric and trims, develops the pattern, makes the samples and delivers a finished product. In the cut and sew model you bring the cloth and the specification while the factory brings the machines and the hands. CMT is the same idea under a different name — cut, make and trim — and "trim" there means the small components, not the finishing work.

A first collection usually starts here if you already have fabric relationships, and full package if you do not. The cost difference is not the sewing. It is who carries the risk of a fabric that arrives late, wrong or short, and our guide to CMT and full package models walks through how to choose between them.

How to brief a factory without a tech pack

You do not need a formal tech pack to start the conversation. You need a reference garment or clear photos, a fabric with a name and a weight, a size you can wear, and the measurements that matter to you. We build the digital pattern, the size grading and the spec sheet from those four things.

What we cannot work from is a mood board and a hope. Send the reference and we will tell you within 48 hours whether the shape is achievable in your fabric at your quantity. If it is not, we say so before sampling rather than after.

Grading is worth agreeing on early. If your sizes run XS to XL, we need to know whether you want the same silhouette scaled up or a shape adjusted for a fuller frame. The second option costs a little more in pattern time and saves a lot of returns.

Three questions to ask before you send fabric to any factory

Ask what the marker yield was on the last similar style. Ask who pays for fabric damaged in cutting. Ask what happens to your leftover cloth, labels and zips. A factory that answers all three in writing is one you can hand 200 metres of viscose to without losing sleep.

We answer those three inside the quotation itself, and our FAQ page carries the fuller checklist. To see the sequence end to end, our production process page runs from tech pack to delivery, and the case studies page shows delivered orders at this scale.

If you already know you want the service and not the sourcing, our cut and sew program sets out what we take on from 30 pieces per style. One note on scope: we do not hand-bead, and we do not run wool coats or knitwear. Saying that on the first call saves us both a month, and our Fashion Pains Solver — in training and coming online soon — exists to name that kind of mismatch before anyone pays for samples.

"High end" gets used as a price tag in this industry. A high end clothing manufacturer is not one that charges more. It is one that holds standards you can check on a garment in your hand, and that is a different thing from a nicer website. Here are the eight standards we hold on every order, including the 30-piece runs.

Does high end production actually cost more at 30 pieces? Some of it does and some of it is free. Fabric inspection and a sealed sample cost hours, not dollars, and a second mock-up costs one sample. The expensive habits are the ones you would keep anyway, which is why we build them into the quote instead of selling them as upgrades.

What high end means in a production order

Price is the worst signal a buyer can use. Two factories can quote the same $68 for a viscose midi dress, and only one of them will measure the finished garment against the spec sheet. The difference shows up in week four, when the collar sits differently on two of your thirty pieces.

We define the standard by what is verified rather than what is claimed. Every claim on this page is backed by a photo, a record or a sample you can hold, and you are welcome to ask for any of them before you pay a deposit.

The eight standards we hold on every order

These are not aspirations. They are steps on a checklist, and each one leaves a record.

• Fabric inspected on arrival. Every roll is checked against the approved swatch for shade, weight and width before it is cut. A roll reading 0.8 Delta E off the standard is held, not hidden.
• Shrinkage tested before cutting. A 10 cm square is washed and measured. If a viscose moves more than 3 percent, the pattern is adjusted first.
• A fit block before a style. Your measurements are cut into a block that then gets graded, so a size 14 does not inherit a size 8 silhouette.
• Two mock-up rounds. Round one proves the shape. Round two proves the fabric, the finish and the measurements on a live fit.
• A sealed sample. The approved garment is signed, dated and kept. Bulk is compared against it, not against anyone's memory.
• Three inspections. Inline after the first pieces, mid-run on the line, and pre-shipment on every piece.
• Mannequin fitting per piece. Each finished garment is fitted and checked for drape, length and symmetry before packing.
• Written approval at every gate. Nothing moves forward on a verbal yes. Every stage is signed off before the next one starts.

A standard only counts if it can fail. When a satin order arrived in two shades last spring, we held the roll, sent the client the Delta E readings and cut only the batch that matched. The order shipped four days late. The alternative was thirty garments that looked mismatched under shop lighting, and no client forgives that twice.

Where cheap production actually shows up on a garment

Not in the seams you can see, but in the ones you cannot. A skipped interlining makes a lapel collapse after two wears. A seam that was never pressed shows as a ridge down the front. A care label sewn into the wrong position becomes a complaint you cannot answer.

Pressing is the cheapest step to skip and the easiest to see. A properly pressed garment has crisp edges and flat seams, while one that was rushed shows a ridge down the front and a lapel that curls. Put your approved sample and a bulk piece side by side under the same light and the answer takes ten seconds.

These are all finishing faults, and finishing is the first place a rushed order cuts. If a garment feels thinner than the sample you approved, it usually is — and it is worth putting both on a table before you accept the balance.

A high price and a quality garment are not the same thing

A $400 garment can be made badly and a $120 garment can be made properly. What changes with price is usually the cloth, the trim and the labour minutes, not the discipline. Discipline is a decision, and it costs less than most brands assume.

That is why we publish an inspection schedule rather than a quality promise. Our high end clothing program lists the same eight standards with the record we keep for each one, and you can request the last three months of them before you order.

A useful test is to ask what a factory checks and when. If the only answer is "before shipping", the order had two chances to go wrong that nobody looked at.

What we refuse to do, at any quantity

We do not run wool coats, knitwear or hand-beaded pieces, and we say so on the first call rather than in week six. We also do not accept an order we cannot inspect properly. If the calendar cannot fit three inspection passes, we move the date instead of the standard.

Saying no early is cheaper than apologising late. Two orders in the past year went to other factories because our floor was full, and both clients came back the following season. That tells us the honesty was worth more than the booking.

If you are comparing factories, our FAQ page collects the questions buyers ask us most, and the case studies page shows what these standards look like on delivered orders. Our production process page maps the three stages and the gates between them. A1, the design illustration agent we are training now and bringing online soon, is being built to the same idea — production checks attached to every design it produces.

Every small batch clothing manufacturer charges more per piece below 50 units, and we are one of them. The question worth asking is not whether the premium exists but where it goes. Here is the arithmetic on a real 30-piece viscose dress, including the costs that never make it onto a quote.

The short answer: the sewing does not get expensive at 30. The setup, the changeover and the waste do. Once those three lines are visible, the premium stops looking like a markup and starts looking like a floor you can plan around.

The fixed cost that never shrinks

Pattern making, grading, two mock-ups, one final sample and line setup cost the same at 30 pieces as at 500. Call that block F. On a lined viscose dress with a zip and a facing, F runs about $1,100 with us — pattern $260, grading $180, mock-ups $340, final sample $180, setup and marker $140.

Divide $1,100 by 30 and you get $36.67 per piece. Divide it by 500 and you get $2.20. Fabric, thread and sewing minutes barely move between the two quantities. That one division is most of the gap, and it explains why a 30-piece quote and a 500-piece quote are not the same garment at different prices.

Where a 30-piece run's money actually goes

Take one style, 30 pieces, viscose at $4.20 a metre and 2.1 metres per dress including the marker allowance. Fabric comes to $264.60. Trims — invisible zip, hook and eye, thread, main label, care label, hangtag — come to $186 for the run, because trim minimums mean you buy for 50 and use 30.

Cutting and sewing runs $22 per piece, so $660. Finishing and pressing is $4 per piece, $120. Three inspections add $3.50 per piece, $105. The sample and setup block is the $1,100 above.

Total outlay is about $2,436, or $81.20 per piece.

Now the line most quotes leave out: freight and duty. A 12 kg air shipment to the US runs $180 to $260, and duty lands around 12 to 16 percent of the declared value depending on the HS code. Call it $380 landed. Your real cost is $2,816, or $93.87 per piece.

Then divide by what you actually sell rather than what you make. If you retail at $290 and sell 24 of the 30, your cost per garment sold is $117.33. That is the number to plan around, not the $81 on the quote.

The cost brands forget: deadstock, samples and cash

You bought fabric for more pieces than you made. On this run, 200 metres yields about 180 garments of this style, so 30 pieces used 63 metres and the rest sits on a shelf. If you never cut a second style from that roll, the shelf is a cost you paid for.

Sample costs are usually quoted separately and deducted from bulk once you order, which is fair. They still leave your account first. Two mock-ups and a final sample on this style were $520 spent before the bulk order existed, and that is cash flow, not unit price.

Payment terms deserve their own line. We ask 30 percent at order confirmation and 70 percent before shipping, so on this run you are out $845 before anything is cut and the balance before it leaves Guangzhou. If your retail sales land after delivery, plan for four to six weeks with money sitting in stock rather than in your account.

Two smaller lines matter more than they look. Packing materials and cartons run about $2.20 per piece at this size, and a carton standard your stockist will accept costs more than the cheapest one. An international transfer fee is usually $30 to $50, paid by whoever the agreement says.

When 30 is genuinely the cheaper number

When the style is unproven. Thirty pieces lets you catch a grading fault or a shade mismatch across one roll instead of across 500 garments. Fixing a grade and reordering costs you a week, while fixing 500 costs you the season.

When cash matters more than margin. Thirty pieces at a $94 landed cost ties up $2,816. Five hundred is roughly $47,000 before a single garment sells. If your sell-through is uncertain, the smaller number buys you information, and information is cheaper than inventory.

Our comparison of MOQ 30 and 500 works the per-piece side of this in more detail, including where the premium disappears.

How to lower the per-piece cost without lowering the standard

Four moves, in order of effect. Raise the quantity to 50 per style, which cuts the fixed block per piece from $36.67 to $22. Use one fabric across two styles so the roll is fully consumed.

Cut the marker tight, because an 8 percent yield gain on this order is $21. Move the shipment to sea freight if you have 30 days of slack, which saves roughly $120 here but adds three weeks you may not have.

What we will not do is remove inspection passes to hit a price. If a quote looks unusually low at 30 pieces, ask which of the three inspections it dropped, and ask for the fabric inspection record. Those two questions settle most price gaps in one email.

If you are still choosing a quantity, our small batch program sets out what we run from 30 pieces, and our FAQ page answers the cost questions we get most often. The case studies page has delivered orders with their numbers attached, and our production process page shows the three stages these costs map to. The Fashion Pains Solver we have in training and are bringing online soon is meant to do one job: name the pain before you spend money on samples.

Two things we do not make, so you can rule us out early: wool coats and knitwear. Our fabrics are viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting. Saying that on the first call saves everyone a month.

Most brands who email a private label clothing manufacturer use the wrong word for what they want, and it costs them about a month. Private label means your name goes on a design the factory already has; white label means you buy a finished design exactly as it is; OEM means the factory builds a pattern that did not exist before. At 50 pieces, one of those three is almost always the right buy.

The short answer: if you already have a fit you can measure and fabric you own, you want OEM. If you need stock on a rail in six weeks, white label. Private label sits between them, with your label and trims on an existing block cut in your fabric. Getting this wrong is the most common reason a first order drifts three weeks late.

What each model covers on a private label clothing order

Private label: you choose a block the factory already grades, then specify fabric, colour, label and trims. The pattern is not yours. Another brand can order the same block next month, and sometimes does.

White label: you buy the garment as the factory makes it, often with nothing changed but the label. It is the fastest and cheapest option, and the least yours. Two shops on the same street can sell the identical dress.

OEM: the factory drafts a pattern from your measurements or your reference sample. This is what most designers mean when they say private label, and it is the only model that gives you a fit identity you own.

Full package sits on top of OEM. The factory sources fabric and trims instead of you supplying them, which moves the sourcing risk to the factory and adds roughly 8 to 14 percent to the unit price.

White label vs private label: the difference buyers miss

White label is a product decision. Private label is a branding decision. You can sew your label into a white label garment, which is why the two get confused, but you cannot change its fit, its fabric or its construction.

The test takes one question. Ask the factory whether another client can order the same garment next month. If the answer is yes, it is white label or a stock block. If the answer is no, you have paid for a pattern and you own something.

On 50-piece runs we see white label used for test drops and accessories, and private label or OEM used for the one piece a brand wants to be known for. That split is usually right.

Where OEM fits a 50-piece run

OEM at 50 pieces is normal work for us. Pattern making, grading, two mock-ups and a final sample are the same effort at 50 as at 500, which is why the unit price is higher at the small end.

That block runs about $1,100 on a lined dress. Split across 50 pieces it is $22 each. Across 30 pieces it is $36.67. The pattern is also yours, so your second order carries no pattern cost at all.

We got this wrong in 2019. We took an OEM brief at 30 pieces and buried the pattern cost inside the unit price so the quote looked cheaper. The brand then changed the neckline twice and we absorbed the difference. Since then the pattern line is quoted on its own, in writing, before anything is cut.

Cost and lead time at 50 pieces, by model

White label, 50 viscose dresses: roughly $28 to $40 per piece, four to six weeks, no pattern cost, and you accept the fit as it comes.

Private label on a stock block, 50 pieces: roughly $46 to $60 per piece, six to eight weeks, because you add fabric selection, a label run and one fitting round.

OEM, 50 pieces: roughly $68 to $92 per piece including the pattern and two mock-ups, eight to ten weeks. The second order falls to about $52 to $70 because the pattern already exists.

Those are Guangzhou ranges for viscose, satin, linen, cotton-linen, rayon and TR suiting, not quotes. Fabric choice moves the number more than the model does. A satin upgrade alone can add $9 to $13 per dress.

How to choose in one page

Answer four questions. Do you have a fit you can measure, do you already own the fabric, does the design have to be yours alone, and can you wait eight weeks?

Yes to the first two and you can wait: OEM. No fit and no fabric but you need stock now: white label. Your own fabric and label but no pattern budget: private label on a stock block.

If you cannot answer the fabric question, settle that before you choose a model. Fabric is 30 to 45 percent of garment cost, and it decides your lead time more than the factory does.

What we will not do

We will not call a stock block OEM to make an order sound bespoke. We will not take payment for a pattern and then run it for another client.

Two things we do not make, so you can rule us out early: wool coats and knitwear. Our fabrics are viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting.

If private label is the model you want, our private label program sets out what we run from 30 pieces. The FAQ page answers the questions we get most often, and the case studies page lists delivered orders with their numbers attached. Our production process page shows the three stages every model runs through. The Fashion Pains Solver we have in training and are bringing online soon exists to name the pain before you spend money on samples.

Every quotation says high quality, and the phrase on its own tells you nothing you can check. A high end clothing manufacturer should be able to put a number next to it: stitch density, measurement tolerance, shade rating, seam strength. These are the seven specs we write into every order, including the 30-piece runs, and you can ask any factory for them this week.

Why it matters at your size: a 50-piece order has no room for a quiet downgrade. There is no second container to hide a bad roll in, and one shade mismatch across 50 pieces is your whole season. Numbers in the order are how you find out before the garments are sewn, not after.

Stitch density and seam strength: the first two numbers

Stitch density is stitches per inch, or SPI. On a woven seam we run 10 to 12 SPI. On fine viscose and satin we run 12 to 14, because fewer stitches on a slippery fabric pull open at the seam.

Seam strength follows. Standard seam allowance is 1 cm, but at stress points — armhole, crotch, side seam on a fitted dress — it should be 1.5 cm with a backstitch at both ends. If a factory cannot answer this in writing, they are not measuring it.

Fabric tolerance and shade rating on a delivered roll

Ask for fabric weight tolerance of plus or minus 5 percent against the approved swatch, and a shade rating of 4 to 5 on the grey scale for colour fastness. Both are standard mill tests, both produce a number, and both take under an hour to run.

Then ask what happens if the delivered roll fails. Ours is simple: the roll goes back and the schedule moves. We have held a satin delivery four days over a double shade on one roll, and we would rather explain four days than 50 dresses that do not match each other.

Two more numbers belong on the same sheet. Shrinkage after one wash should stay inside 3 percent for viscose and 2 percent for cotton-linen. Colour fastness to rubbing should hold at grade 4 for dark shades, which is where cheap dyeing shows up first.

Both are measured on the swatch you approve, so ask for the figures at approval rather than after delivery. A mill that runs these tests has the numbers ready and will send them the same day.

Measurement tolerance and symmetry across a size run

Critical points — bust, waist, hip, shoulder, garment length — should hold plus or minus 0.5 cm against the approved size chart. Non-critical points can hold plus or minus 1 cm. Anything looser is a different garment at size 8 and at size 14.

Symmetry is the check nobody asks for. Left and right sides should match within 3 mm on collar points, pocket placement and sleeve length. On a 50-piece run we measure every garment on a dress form, piece by piece, and the asymmetry we catch most often is a collar point.

Finishing, pressing and the inside of the garment

Finishing is where cheap production shows first. Look for skipped stitches, loose thread ends, uneven topstitching and shine marks from a hot iron on dark fabric. Pressing should happen in stages, not once at the end.

Then turn the garment inside out. Interlining at the collar and front should be fused evenly with no bubbles, and seam allowances should be finished rather than raw and fraying. The inside is the honest part of a garment.

Ask for the finishing checklist with the order, not a verbal promise. Ours has 14 lines, and the three that fail most often anywhere are thread trimming, iron shine on dark fabric, and a crease pressed in where the garment was folded while still damp.

Two more specs: hardware and thread

Hardware gets a pull test. A zip slider should survive 6 kgf without deforming, and a shank button should hold the same. Ask which hardware is used by name, because a substitution here stays invisible until a customer returns the garment.

Thread should be matched to fabric composition and colour, and core-spun on stress seams. On satin and TR suiting we use a fine poly thread with a short stitch. On linen we use a slightly heavier one.

If you use a metal zip or a metal trim, ask for a nickel-release statement. It is a two-line document, it costs nothing to produce, and it decides whether a shipment clears a European stockist's intake check or gets sent back to you.

What to do when a factory says all seven are standard

Ask for the last inspection record on a similar garment. A factory that holds these standards has a filled-in form with numbers, dates and a name on it. A factory that does not will answer with adjectives.

Then ask one more question: which of the seven did you fail last year, and what changed afterwards? Nobody passes every line on every order. The answer tells you more about a factory than the quotation does.

How to ask for these without a fight

Put the seven specs in one table and attach it to the tech pack as an inspection standard. Then ask the factory to confirm each line before you pay the deposit. A factory that runs a real inspection process will answer within a day.

We have been on the wrong side of this. In 2021 we shipped a rayon order where the spec said "match the approved sample" and nothing about shade rating. The rolls arrived two tones apart and we spent three weeks arguing about what "match" meant. Write the number, not the word.

Our high end production page lists the standards we hold on every order, including the small runs. The FAQ page covers the questions we get most often, and the case studies page shows delivered orders with their numbers attached. Our production process page maps these checks to the three stages of an order. The A1 design agent we have in training and are bringing online soon is being built to flag a spec gap before a sample is cut.

Fabric is 30 to 45 percent of your garment cost, and it is the part a small batch clothing manufacturer cannot fix for you. Mills sell by the roll, not by the dress, and a brand with a 50-piece order needs about 120 metres, not 1,000. Here is how to buy at that size without paying a penalty for it.

The short answer: buy stock fabric from a mill or a trading house, accept a roll minimum, and skip custom dyeing until your reorder. Custom dyeing at 100 metres does not exist at a fair price. If your design truly needs a colour that is not on the market, plan for 300 metres minimum and a five-week lead time.

Why mill minimums exist and what they really are

A mill does not sell cloth. It sells machine time. Setting up a loom or a dye bath takes hours, and the mill charges for the run rather than the metre, so 100 metres costs almost as much to produce as 600.

Real minimums in Guangzhou today: stock fabric, no minimum, sold by the roll or cut by the metre. Dyed-to-order in a mill's existing quality, 300 metres per colour, sometimes 600. Custom weave or knit, 800 to 1,500 metres. Custom print, 500 metres per design.

Stock fabric vs dyed-to-order: three fabric minimums

Stock fabric is cloth the mill already runs and keeps. You choose from a card, pay by the metre, and can often take 100 to 150 metres next week. The trade-off is that you do not pick the exact shade, and another brand can buy your fabric the week after.

Dyed-to-order means the mill dyes a base quality to your colour. You get the shade you asked for and a lab dip record on file, but you commit to 300 metres minimum and three to five weeks.

Custom fabric means the construction is yours, whether a specific weave, weight or blend. This is where the 800-metre minimum lives, and it belongs on a second order, not a first one.

How to buy 100 metres without paying mill prices

Three routes work. Buy a full roll of stock fabric, usually 50 to 100 metres, and take the whole roll so you get one dye lot. Or buy from a trading house that stocks mill qualities and sells cut lengths at a small premium, often 6 to 10 percent.

Or buy from the fabric market. In Guangzhou, a 100-metre order at the market runs about $4.20 a metre for a mid-weight viscose, against $3.80 at 300 metres and $3.40 at 600. The market also lets you buy 20 metres in three colours to test, which is a real advantage at your size.

One warning: the same "viscose" at $2.80 and at $4.20 is not the same cloth. Check weight in GSM, shrinkage after one wash, and whether the shade holds. Ask for a half-metre swatch cut from the exact roll before you commit to 100 metres.

What a sampling charge is and when it comes back

Mills charge for a lab dip or a sample metre, usually $80 to $200, and most credit it back if you order above 300 metres. Below that, the charge stays. Budget it as a cost, not a deposit.

Ask for the swatch to be cut from the production roll and kept on file with a lot number. If the bulk does not match the swatch, that lot number is your evidence. Without it, the conversation becomes an opinion, and opinions cost weeks.

Deadstock and job lots: the honest trade-offs

Deadstock is leftover cloth from another brand's run, sold cheap. It can be excellent, a $9 silk-blend at $3.50, and it arrives with three problems: you may only get 80 metres, the shade may drift across the roll, and there is no reorder, ever.

We made this mistake with a client in 2022. They bought 60 metres of deadstock linen to save $1.10 a metre. The shade ran two tones from one end of the roll to the other and the last 12 garments did not match the first 20. We should have inspected the roll before it shipped and told them to walk away.

Deadstock works when the style is a limited drop, the colour is close to something you can live with, and you never plan to reorder. It fails the moment a stockist asks for the same dress in six weeks.

What to do if you cannot find your fabric at 100 metres

Change the fabric, not the factory. There are thousands of stock qualities in viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting. A 5 percent weight change or a half-shade difference is usually invisible to a customer and very visible on your invoice.

Or split the order. Make 50 pieces in one stock colour now, then place the 300-metre dyed-to-order in the colour you actually wanted once the style has sold. You pay the setup twice, but you pay it with money instead of risk.

Start with our small batch program if you want the quantity side worked out, and our cut and sew service page if you plan to supply your own cloth. The FAQ page answers the sourcing questions we get most, and the case studies page lists orders with the fabric and metre counts behind them. Our production process page shows where fabric approval sits in the three stages. The Fashion Pains Solver we are bringing online soon is built to catch a sourcing problem before it becomes a production problem.

Ask an ethical small batch clothing manufacturer for proof of how your clothes were made and you usually get two things: a photo of a clean sewing floor and a certificate logo on a homepage. Neither tells you who sewed your garment, how many hours they worked last week, or what happens when the order runs late. Six documents answer that, and you can check all of them in about a week.

What should you ask for first? Four items, in one email: the address of the workshop that will cut and sew your order, last month's overtime record for that line, the wage payment cycle, and the mill report for the fabric you chose. Ask before the deposit, ask in writing, and ask who signs each one.

Who sews your small batch order: in-house floor vs subcontracted work

Start with one question. Does the workshop that quoted you also cut and sew the garment? Many Guangzhou trading offices quote, then pass the work to a subcontractor you never meet. There is nothing dishonest about that model, but it changes who is accountable when a seam fails.

Ask for the workshop address and the number of machines on the line. Then ask whether your style can be sewn there. We run our own floor in Panyu with a team that has been with us since 2008, and we say so when a style sits outside what we sew.

If a supplier will not give you the address, treat it as a pricing question rather than a moral one. You are not being lied to about ethics. You are being quoted by a middle layer that adds 8 to 15 percent and answers to nobody at the machine.

Overtime and wage records: what a Guangzhou factory can show you

Long hours are the part of garment production brands ask about last and should ask about first. Request the attendance and overtime sheet for last month on the line that would run your order. A real factory has one, with names and hours, usually on paper.

Read the hours, not the summary. A floor running 55 to 60 hours in peak season is normal in Guangdong, and it is a fact you want before you print a delivery date in a lookbook.

Then ask how overtime is paid, and whether it lands on the same cycle as the base wage. A factory that answers without checking with anyone has a payroll system. One that says "we treat our workers well" does not.

We hold our own records and send the overtime sheet on request. We do not send individual wage amounts, because that is a worker's information, not a sales document.

Fabric certificates and dye lots: how to verify a mill report

Every mill report carries a number and an issuing body. Copy the number, open the issuing body's site, and check it. Two minutes, and it separates a real test report from a logo file.

For fabric, the two documents that matter are a fibre composition report and a colour fastness report. If you sell into the EU, a REACH statement on restricted substances is worth more than any eco label, because it is the one your buyer's compliance team will ask for.

Ask which certificate applies to which roll, not to the factory. Reports attach to batches. A valid report from last spring tells you nothing about the cloth arriving next month.

We do not hold GOTS or OEKO-TEX certification under our own name. We request mill reports on the fabric you choose and pass them on, and we say plainly when a report does not exist.

Offcuts, deadstock and the waste question at 30 pieces

Small runs create more waste per garment than large ones. A marker for 30 pieces leaves a bigger offcut percentage than a marker for 500, and nobody in this industry likes saying so out loud.

Ask what happens to the offcut. Ours goes to a textile recycler in Panyu, with a small amount returned to sample work. Ask the same question about deadstock, because a factory holding 260 metres of your dyed fabric after a cancelled second order is a real cost.

Write down who owns leftover fabric before the first cut. We had a brand cancel a second run in 2022 and leave us with 260 metres of dyed viscose. That clause is in our quotation now, in writing, before any cutting starts.

The two things we cannot prove

We cannot prove a wage figure ourselves, because no factory self-report is an audit. A third-party audit is the only document that means anything, and it costs money a 30-piece order rarely carries.

We also cannot prove a fabric is traceable to a farm. Traceability claims deeper than the mill usually stop at a trading company, and any factory repeating them is passing on a supplier's marketing.

What we can give you is a named team, our own records, and a straight answer when a document does not exist. That is a smaller claim than most factory websites make, and it is one we can stand behind on a Tuesday.

How to run this check in one week

Send four requests in a single email: workshop address, last month's overtime sheet, the wage cycle, and the mill report for your chosen fabric. Give a date. A factory with records replies in a day or two.

Then ask the same four questions of the next factory you contact. The comparison is worth more than either answer, because the gap between the two replies tells you which one is organised.

Our small batch program sets out what we take from 30 pieces and what we ask you to supply. The FAQ page covers the questions we get most often, and the case studies page lists delivered orders with their numbers attached. Our production process page shows the three stages and the written check at each gate. The Fashion Pains Solver we are bringing online exists to name the cost of a bad sourcing decision before you pay a deposit.

A private label clothing manufacturer will quote you a garment price and leave out the four lines that decide whether 30 pieces is even possible: woven labels, care labels, hangtags and polybags. Each one carries its own minimum order quantity, and the minimums are usually larger than your whole run. On a 30-piece order you buy the minimum and store the rest.

What do those minimums look like in practice? Woven labels commonly start at 1,000 pieces per design, printed care labels at 500 to 1,000, and hangtags at 500 to 1,000 plus a one-off plate charge. Custom printed polybags start at 5,000. Added together, trims can reach $7 to $14 per garment on a run of 30.

The reason is structural rather than greedy. Labels, tags and bags are printed or woven on a separate line with its own setup, and the supplier behind that line has no reason to run it for 30 units. The minimum belongs to the printer, not to your factory, and your factory cannot negotiate it away.

Woven labels and care labels: the 1,000-piece problem

A woven label runs $0.06 to $0.12 each at 1,000 pieces, so $60 to $120 buys a lifetime supply for a 30-piece order. Divide that by the 30 garments you actually use and the label costs $2 to $4 per piece.

Care labels are cheaper per unit but have their own rule. They must state fibre composition and care symbols, and in the EU they must appear in the language of the market you sell in. A care label written for the US will not pass an Italian stockist's intake check.

Two ways down. Use a printed label at a 500 minimum instead of woven, which cuts the label cost by roughly half. Or buy stock labels in a neutral colour and overprint your name in the centre, which works while you are still testing a style.

Hangtags, barcodes and retail packaging minimums

Hangtags usually start at 500 to 1,000 pieces, run $0.08 to $0.20 each, and carry a one-off plate or die charge of $30 to $80. That plate charge does not shrink with your order size, which is why a 30-piece run pays the same setup as a 3,000-piece run.

If you sell to a stockist you will also need a barcode. A UPC or EAN comes from GS1 in your own country, not from the factory, and it is a yearly cost rather than a one-off. Check the current fee where you are registered, because we cannot issue one for you.

Polybags are the line most brands forget entirely. Plain stock bags come in the hundreds. A custom printed bag with your logo starts around 5,000, and several markets now restrict single-use packaging on top of that.

What 30 pieces actually costs in trims

Here is a real 30-piece breakdown for a viscose dress: 1,000 woven labels at $0.09 is $90, 500 printed care labels at $0.04 is $20, 500 hangtags at $0.12 plus a $45 plate is $105, and plain polybags at $0.06 for 100 is $6. Total $221.

Across 30 garments that is $7.37 per dress before you add thread, buttons or a zip. A brand expecting $3 of trims per garment is off by more than double, and the gap is entirely minimum quantities rather than quality.

The garment price did not change. The trims did, and they were never inside the garment quote, because the factory does not buy your labels.

How to cut trim cost without cheapening the garment

Use one hangtag design across a whole capsule instead of one per style. The plate charge is paid once, and the hangtag stops being a per-style cost.

Buy care labels as a single stock roll and overprint, and keep the fibre percentages in a template you reuse. Most brands change one fabric and then reorder 1,000 care labels they did not need to reorder.

Buy polybags plain and add a sticker. It reads as intentional in a small brand's packaging, and it removes a 5,000-piece minimum from your first order.

None of these changes touch stitch density, seam allowance or fabric. Trim savings come from volume and design discipline, not from downgrading the garment.

Where we got this wrong

In 2018 a brand ordered 30 silk dresses and we quoted trims at roughly $3 a garment. The invoice landed near $11 because we had forgotten the hangtag plate charge and assumed the label minimum was 300, not 1,000. We absorbed it and rewrote the template.

Since then trims are quoted as their own block, separate from the garment, with the minimum printed next to each line. A brand can see exactly what they are buying 1,000 of, and decide whether to store it or split the cost with a second style.

What to send with a private label order

Send four files: label artwork as vector, care text with fibre percentages for each market, a hangtag file, and your barcode numbers if you have them. If the artwork does not exist yet, tell us, and we will quote trims as an estimate and mark the lines that will move.

Then decide the storage question early. Four boxes of labels in your apartment is a normal first-order outcome, and it is cheaper than paying the same minimum twice.

Our private label program covers what we take from 30 pieces and which parts of the garment are yours to specify. The FAQ page answers the questions we get most often, and the case studies page lists delivered orders with real numbers. Our production process page maps trims approval into the first stage, before sampling starts. The A1 design agent we are bringing online is being built to flag a missing care-label text before a sample is cut.

Ask a high end clothing manufacturer for 30 evening gowns and most will quote 300. The reason is rarely the sewing. It is the lining, the structure, the length and the fabric minimum: four decisions that get expensive at small quantity, and four you can settle before you ask for a price.

What does a 30-gown run actually need? One fabric decision, one lining decision, one structure decision, one length decision. Fix those four and the number usually comes down by a quarter to a third without changing how the gown looks on a body.

Lining choice: silk, cupro or viscose on a 30-gown run

Lining is the first lever and the one brands resist. Silk lining runs $9 to $14 a metre, cupro $6 to $9, and a good viscose or polyester lining $3.50 to $5.50. A floor-length gown takes 1.6 to 2.2 metres of lining.

Moving from silk to cupro saves roughly $5 to $8 a gown and keeps the breathability and the drape. Moving to viscose saves more, and it is a real downgrade on a garment a customer wears against bare skin for four hours.

Our default on evening wear is cupro. It presses cleanly, it does not squeak under satin, and at 30 pieces the saving pays for one mock-up round.

Structure and boning: what a small run can and cannot do

Boned bodices are normal work for us at 30 pieces. Spiral steel in a channel is not a minimum-quantity item: the channel is sewn, the boning comes by the roll, and the cost is labour rather than material.

What a small run cannot do is hand beadwork. We do not offer hand beading at any quantity, because the labour is unpredictable and a 30-piece order has no way to absorb a 40-hour overrun. Machine embroidery we do run, and it is quoted per 1,000 stitches so you can see the cost before you commit.

If your design depends on dense hand embellishment, we are the wrong factory. Say so on the first call and save us both a month.

Length and hem: where an evening gown eats fabric

Length is the quiet cost. A floor-length gown cut from 140 cm wide fabric needs 2.6 to 3.2 metres per garment once you allow for nap, matching and the hem. A midi needs 1.9 to 2.3 metres.

That 0.8 metre difference is $3 to $6 per gown in fabric and a visible change in how much cloth you must buy. Across 30 gowns it also decides whether you sit under a 100-metre order or over it.

Ask for the marker before you approve the length. We send the marker lay for approval on evening wear, because a 4 cm hem change can move the fabric requirement by 12 metres across a run.

Fabric minimums and dye lots at 30 pieces

Stock satin, crepe and viscose have no minimum. Dyed-to-order starts at 300 metres per colour, which is ten times what 30 gowns need. That single number is why small evening wear brands should design around stock cloth first.

If you must dye, use one colour across two styles. Splitting 300 metres between a gown and a slip dress brings the deadstock down and gives you a second SKU from the same dye lot.

Ask for the dye lot on the invoice. We held a satin delivery four days in 2021 over a double shade on one roll, and four days beats 30 gowns that do not match each other under stage light.

What we refuse to make

Two categories, so you can rule us out early: wool outerwear and knitwear. Hand beading we also decline at any volume. Our fabrics are viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting.

We do not take single-piece couture either. The three-stage process we run depends on a bulk run behind the sample, and a one-off gown is a different trade with different skills.

Cost at 30 gowns, line by line

A viscose or crepe column gown with cupro lining and no boning: roughly $52 to $70 per piece at 30 units, 8 to 10 weeks including two mock-ups. A satin gown with a boned bodice and a full lining: roughly $78 to $110.

At 300 pieces the same satin gown falls to about $58 to $80, and the gap is setup rather than sewing. Pattern making, grading, two mock-ups and a final sample cost the same at 30 as at 300.

We got this wrong once. In 2020 we ran a 30-piece gown order and did not test the boning under the finished lining weight until the second mock-up. Six gowns had to be re-cut. Structure now gets tested on mock-up one, every time, before the lining is attached.

How to brief an evening wear run

Send the reference, the target retail price, the market, and the four decisions you have already made. If the lining is not chosen yet, say so, and we will quote both options side by side instead of picking one for you.

Then ask for the marker and the trims minimums before you approve the sample. Both are cheaper to change on paper than on a cutting table.

Our high end production page lists the standards we hold on every order, including 30-piece runs. The FAQ page answers the questions we get most often, and the case studies page shows delivered evening wear orders with their numbers attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to catch a structure or lining mismatch before a sample is cut.

Perfect fit is the phrase every brand uses and almost nobody defines. We are a high end clothing manufacturer in Guangzhou, and we publish five numbers that turn it into something you can check: base size, fit model measurements, tolerance, mock-up rounds and grade rules.

What should you ask a factory to prove fit? Those five things, in writing: which base size the pattern is drafted on, who the fit model is and what she measures, the tolerance you accept at each critical point, how many mock-up rounds are included, and how the grade rule moves between sizes. A factory that cannot answer them is selling you a sample, not a production standard.

Here is how we set ours, including the year we got it wrong.

Base size and fit model: the two numbers behind every pattern

We draft every women's pattern on a base size 8, which is EU 38 and IT 42. The base size is the size the pattern is drawn at, and every other size is derived from it by the grade rule. Draft on a size 4 and grade up to a 16, and the larger sizes carry the accumulated error of six steps. When your range is wide, we move the base to the middle of the run instead of the small end.

The fit model is the second number. Ours measures 86 cm bust, 68 cm waist, 94 cm hip and 168 cm tall, and it is the same person on every sample we fit. We log her measurements, and they do not drift between seasons.

If your customer is 5 cm taller or 8 cm larger through the hip, do not accept our block. Send her measurements and we draft a block on your numbers. That costs about three extra days on the first pattern and nothing after that.

Measurement tolerance: the number that decides a pass or a remake

Tolerance is where a fit standard becomes measurable. Ours is plus or minus 0.5 cm at bust, waist, hip, shoulder and garment length. Non-critical points such as pocket placement hold plus or minus 1 cm, and left and right sides match within 3 mm.

Write the tolerance into the order, not into the conversation. We measure every garment against the approved size chart and record the numbers on a sheet that ships with the goods. On a 30-piece run that is 30 rows of measurements, and it is the fastest way to settle an argument six weeks later.

One thing we will not do is quote a tolerance we cannot hold. On a slippery satin or a fine rayon, plus or minus 0.5 cm at the hem is honest. Promising 2 mm on the same cloth is marketing.

Mock-up rounds and fit approval on a small production run

Two mock-up rounds are included in every order, at any quantity. Round one is fit only: bust, waist, hip, shoulder, armhole, and rise where the style has one. Round two is fit and finish together, including lining, hem and closures.

Round one is fitted on the model and then on a dress form, because a form catches the asymmetry a live body hides. Round two goes back on the model before the sample is sealed. A third round is possible and charged separately, and we say that on the quote rather than in week four.

Grade rules: why a size 8 sample says nothing about size 16

A standard grade on a woven dress moves the bust 5 cm, the waist 4.5 cm and the hip 5 cm per size step. Those steps hold until the hip reaches roughly 110 cm, and past that the step has to widen or the garment pulls across the seat.

Stretch fabric is the worst case. A rule set written for woven cloth graded onto jersey gives you a size 8 that fits and a size 16 that clings. We grade jersey on its own rule set, and we fit the two extreme sizes on the model before the bulk marker is made.

We learned this the expensive way. In 2022 we graded a viscose dress to a size 16 with a standard woven rule and the armhole gaped 2 cm on the model. We redrafted the block and re-cut 14 garments at our own cost. Since then the extreme sizes are cut in the base fabric and fitted before bulk, on every style.

What happens when a fit standard fails on our floor

It fails sometimes. The rule is simple: the run stops at the stage where the fault is found, and the schedule moves.

A fit fault caught at round one costs a pattern change. The same fault caught at final sample costs a re-make. Caught after bulk cutting, it costs fabric. That is why we fit the extremes early, and why a third mock-up round is cheaper than a second bulk run.

Ask any factory what they did the last time a fit standard failed. The answer tells you whether the standard is real.

How to put a fit standard in writing before you pay a deposit

Attach one page to the tech pack: base size, fit model measurements, the tolerance table, the number of mock-up rounds, and the grade steps. Ask the factory to confirm each line before the deposit leaves your account.

Then keep the approved size chart with the order. The chart, not the sample, is what the floor measures against, and it is what an inspector should be handed on arrival.

Our high end production page lists the standards we hold on every order, including 30-piece runs. The FAQ page answers the fit questions we get most often, and the case studies page shows delivered orders with their numbers attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to flag a fit standard that is missing before a pattern is cut.

Starting a clothing brand with 30 pieces per style is the cheapest way to find out whether a design sells. We are a small batch clothing manufacturer in Guangzhou, and most of our first orders come from one-person brands. Thirty pieces is not a compromise on quality. It is a test you can afford to repeat.

How many styles should a first collection have? One to three, not ten. Each style carries its own pattern, its own grade and its own mock-up rounds, so ten styles means ten setup costs paid before a single garment sells. Three styles at 30 pieces is 90 garments and a small online store with something to show.

Below is what 30 pieces actually covers, what it costs, and what has to be ready before you ask for a quote.

What 30 pieces per style actually covers

Our minimum is 30 pieces per style per colour. That is one style, one cloth, one size run. A 30-piece run in five sizes looks like 4 XS, 6 S, 8 M, 7 L and 5 XL, or 8, 8, 8, 6 across four sizes. You set the split, and we will tell you if the smallest size in your curve is too thin to cut sensibly.

What it does not cover is 30 pieces spread across three colours. Three colours from stock fabric is fine, because the cloth already exists. Three colours dyed to order is three 300-metre minimums, which is 900 metres for a 30-piece run and the fastest way to turn a $3,000 order into a $12,000 one.

How many styles should a first collection have?

Start with one style you believe in, or two if you want a second colourway to test. The second order is where a range belongs, because by then you know which shape sold and which size ran out first.

If you must launch with three, keep them on the same fabric. Three styles cut from one cloth share a marker, a dye lot and a fabric minimum, and the third style often costs less than the first two.

One more thing about the shape of a first order. A collection is not a range until you have sold something, and 30 pieces of one style tells you more than 10 pieces of three. Which size ran out first, which colour got asked for, what customers said about the fit. That is the information a second order should be built on, and it costs one style's setup to get.

What a 30-piece first order costs, line by line

On a lined viscose dress we quoted recently, the landed cost at 30 pieces came to $93.87 per piece, against $81.20 before freight and duty. The gap is the small-order premium, and we print it as its own line rather than averaging it into the unit price.

The lines are fabric, cut-make-trim, trims, sampling and freight. Sampling is charged separately and credited back against the bulk order once you place it. Trims are usually the surprise on a first private label order, because labels, care labels and hangtags each carry a minimum bigger than 30 pieces.

So the real question is not the unit price. It is whether 30 pieces can sell at full price. At a retail price of $250, 30 pieces is $7,500 of retail. If two thirds of it sells, you have paid for the next run and learned your size curve.

The 12 to 14 week timeline for a first production run

Tech pack and quotation take one to two weeks, or five to seven days if we build the tech pack for you from reference photos and measurements. Two mock-up rounds take 14 to 21 days. The final sample takes 28 to 42 days, and bulk production runs 6 to 10 weeks.

Add two to four weeks if we are sourcing the fabric, and up to five weeks if the colour is dyed to order. From tech pack to shipping, plan on 12 to 14 weeks for stock cloth and 16 to 18 for a dyed shade. Then add shipping, which is 5 to 7 days by air to the US or the EU.

What to have ready before you ask for a quote

Five things: two or three reference images of the shape from different angles, your target retail price, a size chart or your fit model's measurements, a fabric direction, and your market so the garment is labelled correctly.

You do not need a registered company, an office, a warehouse or a finished tech pack. We quote from a reference and a target price most weeks.

Three mistakes that cost first-time brands money

Ordering ten styles at 30 pieces each. Setup cost is per style, so ten styles is ten times the fixed lines with almost no selling history to justify it.

Choosing a dyed-to-order fabric for a 30-piece run. A 300-metre minimum against a 63-metre need leaves 237 metres of deadstock in your storeroom.

And starting without a size chart. In 2019 we took a 30-piece order from a founder who had no fit sheet, so we cut it on our standard block. The garments fitted our model and not her customers, and 11 of the 30 came back. We now ask for a fit model measurement sheet before we quote, and we draft on the brand's numbers rather than ours.

Our small batch program sets out the quantity side in full. The FAQ page answers the first-order questions we get most, and the case studies page shows delivered runs with their quantities and outcomes attached. Our production process page maps the three stages from tech pack to bulk. The A1 design agent we have in training is being built to turn a reference image into a first tech pack draft.

A blazer manufacturer will usually quote you 300 pieces. We run tailoring at 30, and it changes four things: pattern time, canvas handling, the number of fit checks, and how the cloth gets bought. Below is what actually shifts on a small tailoring run, in numbers.

What does a 30-piece blazer order need that a dress order does not? A fitting model with the canvas basted, three checks that only exist on tailoring, and a fabric plan that survives a 300-metre dye minimum. Get those right and 30 jackets is a normal job for a cut and sew manufacturer with a tailoring floor.

What changes on a blazer run below 50 pieces

A blazer runs 25 to 35 pattern pieces against 6 to 10 for a simple dress, and sewing time per garment is three to four times longer. The setup lines are identical at 30 and at 300, though: pattern, grading, two mock-up rounds, a final sample and a marker.

That is the whole story of the price. Below 50 pieces the setup is divided by 30 instead of 300, and the labour saving from a longer run is smaller than most brands expect, because a blazer is mostly skilled hand time rather than machine speed.

Trims follow the same logic. A jacket needs four to six buttons, two shoulder pads, a length of canvas, fusible interlining, a pocketing cloth and a back lining. Buttons and shoulder pads come by the box rather than by the garment, so a 30-piece run leaves you with spares you will use on the reorder. That is a cost, not a waste, as long as the style repeats.

Pattern pieces and canvas: where tailoring eats time

We build a half-canvas front on TR suiting: a floating chest piece, a taped front edge and a hand-set collar. Full canvas is a different trade with far more hand work, and we quote it separately rather than pretending the two are interchangeable.

Interlining is cut and fused piece by piece, and the collar goes on by hand. On a 30-jacket run that is 30 collars, and it is the slowest single operation on the floor. Nobody speeds it up by making more of them.

What we do not take: wool outerwear, knitwear, and hand beading at any volume. Our cloth list is viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting.

Shoulder, armhole and sleeve pitch: the three fit checks

Tailoring has three checks a dress does not. Shoulder width holds plus or minus 0.5 cm. The armhole has to sit without pulling the chest. And the sleeve pitch has to let the sleeve hang without a twist down the front of the arm.

Round one is fitted on the model with the canvas basted and no lining, because a lining hides a shoulder fault. Round two is the finished jacket on the same model, checked with the arm raised, the arm forward, and the jacket buttoned.

We got this wrong in 2021. We quoted a 30-piece jacket order and did not test sleeve pitch until the finished sample, and eight of the 30 jackets needed their sleeves re-set. That added six days and a lot of unpicking. Pitch is now checked on mock-up one, on the model, on every tailoring run.

TR suiting and lining: fabric numbers for a small tailoring run

TR suiting for tailoring runs 240 to 280 g/m². A single-breasted two-button jacket takes 1.8 to 2.1 metres, and the lining takes 0.9 to 1.1 metres in cupro or viscose. On 30 jackets that is roughly 60 to 65 metres of suiting and 30 metres of lining.

That puts you inside the stock range. Dyed-to-order suiting starts at 300 metres per colour, five times what 30 jackets need, so a first tailoring run belongs on stock cloth and a shade card rather than a custom colour.

Ask for the marker before you approve the sample. A 3 cm change in jacket length moves the fabric requirement by about 6 metres across 30 pieces, which is the difference between one roll and two.

What a 30-piece blazer order costs against 300

A half-canvas TR suiting blazer at 30 pieces runs roughly $78 to $105 per piece FOB, with two mock-up rounds and the final sample included. The same jacket at 300 pieces lands around $58 to $78, and the gap is setup rather than sewing.

Two things move the number more than quantity does. Lining choice, where cupro against a good viscose is $2 to $3 per jacket, and the fabric price step, where 60 metres and 300 metres are not the same rate per metre.

If you are ordering a set, cut the jacket and the trouser on one marker. Thirty jackets and thirty trousers booked as two separate markers means two lays, two fabric bookings and two cutting days for one outfit.

How to brief a small tailoring run

Send the reference, the target retail price, the size run, and the fabric direction if you have one. Tell us whether it is a jacket on its own or a set, because that decides the marker.

Then ask two questions before you pay the deposit: how many pattern pieces the style takes, and how many fit checks the jacket gets. The first tells you what you are paying setup for. The second tells you whether the shoulder will fit a real customer or only the sample.

Our cut and sew service page sets out what a tailoring order includes and what you supply. The FAQ page answers the tailoring questions we get most, and the case studies page lists delivered orders with their quantities and fit outcomes attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to flag a pitch or shoulder fault before a jacket is cut.

Most cut and sew manufacturer quotes start at 300 pieces, and the sewing machine is rarely the reason. The number comes from the cutting room, the fabric minimum, and the way a line is balanced. We run the same service at 30 pieces per style, so here is what actually sets a minimum.

What is the real minimum for cut and sew? For us it is 30 pieces per style per colour, and 30 is not simply a smaller version of 300. Below about 50 pieces you are paying for changeover instead of fabric and minutes, so the quote is built differently. Ask a factory what its minimum is divided by, and you will hear which of those two businesses it is in.

Here are the four things that push the number up, with our own numbers attached.

What a cut and sew minimum actually measures

A minimum is the quantity at which fixed work stops dominating the price. Four blocks sit inside that fixed work: the pattern and grade, the marker, the machine changeover, and the trim minimums.

Changeover is the one that surprises people. Setting a 12-station line for a new style takes about six hours: threads, needles, attachments, tension, and a first-off check at every operation. That is 72 operator-hours before a single saleable garment exists.

At Guangzhou wage rates those 72 hours cost roughly $250 to $350. Divided by 30 pieces that is $8 to $12 a piece. Divided by 300 it is 80 cents to $1.20. Same six hours, ten times the spread, and that arithmetic sits behind most 300-piece minimums.

Marker and lay efficiency: the cutting room math at 30 pieces

A lay is a stack of plies cut in one pass. Thirty pieces across five sizes is one marker, six plies per size, one spread about 6.5 metres long. Two people spread it in 45 minutes and cut it in 20.

So cutting is not the constraint at 30 pieces. A nested five-size marker saves four to five metres against cutting size by size, which is $18 to $25 at $4 a metre. Real money, but not the difference between a $9 and a $20 cut-make price.

The constraint is what you cut. A midi viscose dress takes about 1.9 metres per garment, so 30 pieces needs 57 metres of cloth, and the mill's minimum is where small orders usually die.

Fabric minimums that decide your cut and sew order size

Dyed-to-order cloth starts at 300 metres per colour at most mills. Your 30-piece dress needs 57. The mill's minimum, not the factory's, is what turns 30 pieces into 300, and no amount of negotiating at the sewing floor changes it.

Stock cloth has no minimum. We hold a shade range in viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting, and 30 pieces off a stock roll is an ordinary order rather than a favour.

Custom hardware follows the same rule. A bespoke button or zip pull carries a 1,000-piece minimum, so a first run belongs on a stock trim with your own label sewn inside.

Line balancing: why 30 pieces sits badly on a 12-operator line

A volume line gives each operator one operation. Thirty pieces means each of twelve operators sews two or three garments and then the style is finished, and the line has to be torn down and set again.

Small-run work is organised the other way: four to six machinists, each handling several operations, finishing garments one at a time. Slower per minute, far cheaper per changeover.

Neither model is better. They are priced on different denominators, and a mass cut and sew shop quoting your 30 pieces is usually quoting the changeover it would rather not do.

How to brief a cut and sew order below 50 pieces

Send four things: the reference, the size split, the fabric if you have it, and your target retail price. The size split matters more than people expect, because it decides the marker and the ply count.

Say who supplies the cloth. Cut and sew means we cut, sew, finish and inspect while you supply fabric and trims. Full package means we buy them, and the quote carries a fabric line with its own price risk.

Then ask three questions before the deposit: how many plies will my marker be cut at, how many pattern pieces does the style have, and how many fit checks are included. The answers tell you whether the price is built for your order or borrowed from a bigger one.

What goes wrong on small cut and sew runs

We got one wrong in 2020. A client changed the size split after the marker was drawn, we cut size by size to catch up, lost six metres of cloth, and bought a second roll that arrived from a different dye lot. Nine of the 40 garments were visibly darker. The size split is now locked in writing before a marker is made.

Two other things fail on small runs. Trims arrive after the fabric is cut, so finished garments wait on a label. And a brief arrives with no size chart, so we draft on our block and the fit belongs to us rather than to the brand.

We decline wool outerwear, knitwear and hand beading at any volume, so a brief can be ruled out in one email instead of three samples.

Our cut and sew service page sets out what the service includes and what you supply. The FAQ page answers the minimum questions we get most, and the case studies page lists delivered runs with their quantities and outcomes attached. Our production process page maps the three stages from tech pack to bulk. The A1 design agent we have in training is being built to turn a size split into a marker plan.

A private label clothing manufacturer can only print what you send. Artwork that arrives late, or in the wrong format, is the most common reason a small order slips a week just before cutting.

What do you need to send before labels can be made? Five things: label artwork sized to the fold you want, care and content text in the language of the market you are selling in, a barcode, a hangtag file, and a packing bag specification. Miss one and the labels cannot be ordered, which moves the cutting date rather than the label date.

Here is the order we ask for them, and the two mistakes that cost us real days.

Woven label artwork: size, fold and colour to fix first

A woven label is made on a loom, so the artwork has to be drawn at loom size. Resizing a design afterwards changes the lettering weight and the border, and the second proof is not free.

Three folds cover most orders. Centre fold at 40 by 25 mm, folded to 20 mm, sits in a neckline or a waistband. End fold at 50 by 20 mm folds at both ends. Straight cut at 60 by 15 mm is sewn flat on all four sides.

Two colours on a ground are standard. Every extra colour adds a loom charge, and the first-order loom charge runs $60 to $120 per design, paid once. Reorders on the same artwork pay nothing.

Send vector artwork in AI, EPS or PDF, or a 300 dpi PNG at final size. A phone screenshot will not hold a 1 mm stroke, and we will not guess what a line was supposed to be.

Care and content labels: what each market requires

The EU requires fibre composition as a percentage, in descending order, plus the care symbols, in the language of the member state where the garment is sold. A label in English only does not satisfy that for a French or Italian shipment.

The US requires fibre content, country of origin and care instructions. An RN number comes from the FTC, costs nothing, and belongs to the brand rather than the factory.

We print what you approve, and we will flag a missing country of origin. What we will not do is invent a fibre percentage for a blend we did not buy, or print a care instruction the cloth cannot survive.

That second point is not theoretical. We once printed "dry clean only" on a viscose dress that washes perfectly at 30 degrees, because the copy came from a client's previous supplier. We should have asked, and now we do.

Barcodes, hangtags and polybags: the files nobody sends first

A GTIN comes from GS1, not from a factory. A small GS1 block starts around $250 in the US plus an annual fee, and it is a brand cost you cannot route through production.

Print the barcode black on white at the magnification the symbology requires, with the quiet zones left clear. A barcode in your brand colour looks better on a hangtag and fails at the scanner, which reaches you as a chargeback rather than as a print error.

Polybags sold into the US need a suffocation warning when the opening is over 12.7 cm. We print it into the bag or apply a sticker, and either way it has to be on the spec before the bags are ordered.

Hangtag attachment is a small decision with a labour line behind it. Barbs are standard and cheap; string or a loop costs more in hand time on every single garment.

The bag is a spec too, not a preference. State the size, the thickness in microns, whether it self-seals or carries a header card, and whether the garment is folded flat or goes on a hanger. A wrong bag size means the garment does not fit inside it, and bags run on the same lead time as labels.

When artwork has to be approved, and why it is before cutting

The care label goes into a seam, so it has to be at the factory before the first garment is sewn. Labels are not something you add once the goods are finished.

Woven labels take 10 to 14 days from artwork approval. Printed labels take 5 to 7. Hangtags take 7 to 10 including a proof round. Working backwards, artwork sign-off belongs in Process 1, next to the quotation.

Our calendar puts artwork approval with the tech pack and label delivery before cutting. When a brand approves artwork during sampling instead, the order finishes on time and ships a week late, which is the same thing to a buyer waiting on stock.

One workaround we use on tight orders: approve the care label content first, because it carries the longest compliance risk, and let the hangtag design follow a week later. The care label is the one that can hold a shipment at customs, so it should not wait on a marketing decision.

Five private label artwork mistakes we see every month

Artwork sent at screen size rather than final size, so the fold comes out short. Care labels written in one language for a two-market launch. Barcodes printed in a brand colour. And a neck label with a brand name and no fibre content, which is a legal requirement in the EU.

The fifth one is a hangtag approved after bulk cutting, so finished goods sit in the packing room waiting for a tag. It reads as a small delay and it lands in the week you were planning to ship.

The mistake that cost us most was smaller than any of those. In 2021 a brand sent label artwork as a phone screenshot. We rebuilt it, the fold came out 2 mm short against the loom, and 30 neck labels had to be reprinted.

Eight days and $180 on a 30-piece order. Vector file and a stated fold, every time, before anything is ordered.

Our private label program page sets out what we handle and what you supply. The FAQ page answers the labelling questions we get most, and the case studies page lists delivered orders with their quantities attached. Our production process page maps the three stages from tech pack to bulk. The Fashion Pains Solver we are bringing online is being built to check label content against the destination market before a print order goes out.

A small batch clothing manufacturer uses the word custom differently from a trading company. Custom should mean your pattern, your fit, your cloth and your labels, not a stock style with your tag sewn in.

What does custom actually include at 30 pieces? Your own pattern and grade, a fit model fitted over two mock-up rounds, your choice of cloth from a stock range, and your own neck label, care label and hangtag. What it does not include at 30 pieces is a dyed-to-order cloth or custom hardware, because both carry minimums ten times your run.

Here is where the line sits, and what the custom part costs.

Custom pattern and grade: the line between custom and stock

A new pattern built from your reference images and your fit model's measurements runs $120 to $260, and we credit it against the bulk order once it is placed. Adjusting one of our existing blocks to your measurements runs $40 to $90.

The grade rule is the part people forget. A pattern is one size; a grade is the rule that moves bust, waist and hip between sizes, and it has to be set to your range rather than borrowed from a standard run.

Three questions separate a custom supplier from a relabelled one. What base size is my pattern drafted on? How many pattern pieces does the style have?

Then ask what the hip grade step is. A supplier selling stock with your label cannot answer that one, because the rule was written for a different block on a different body.

Pattern storage matters more than it sounds. We keep the block, the graded nest and the fit comments in one folder against the style number, so a reorder is a cutting ticket rather than a project. A brand that reorders twice a year gets its third run at the lowest price of the three without changing a single spec.

Custom patterns for designer labels ordering small quantities

For a designer, the pattern is the product. That is why we draft on your measurements and keep the block on file with the style number, so a reorder six months later starts from the same pattern instead of from a fresh interpretation.

It also means your first order carries the pattern cost and your second does not. On a 30-piece run a $180 pattern is $6 a garment. On the reorder it is zero, and the same style gets cheaper without anything about the sewing changing.

What we will not do is pretend a five-style first collection is cheap. Five styles is five patterns, five grades and five sets of mock-ups, and that is the number that decides a first order size rather than the piece price.

Custom cloth: why a stock fabric is not a downgrade

We hold a shade range in viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting, sold by the metre from stock with no minimum. A 30-piece midi dress needs about 57 metres, and 57 metres off a stock roll is an ordinary order.

Dyed to order starts at 300 metres per colour at most mills. Buying that minimum for a 30-piece run leaves about 240 metres in your storeroom, which is deadstock with your name on the invoice.

What a custom shade buys you is a colour nobody else is selling this season. If that is the point of the collection, buy it once across two styles so the 300 metres is spread over 60 pieces instead of 30.

Custom labels, care labels and packaging at 30 pieces

Neck label, care label, hangtag and bag can all carry your brand on a 30-piece order. The artwork is what costs, not the quantity, so the design fee is the same at 30 as at 300.

Care label content has to match where you sell: fibre composition as a percentage, country of origin, and the language of the country of sale. For the EU that means the language of the member state, not English by default.

Keep one hangtag design across the season. A second hangtag is a second plate charge and a second proof round for no selling benefit.

What custom work costs at 30 pieces

Here is the one-off side of a custom 30-piece order, in our own numbers. Pattern $180, grade $60, two mock-up rounds included, woven label loom charge $90, care label plate $40. About $370 before a metre of cloth is cut.

Divided by 30 that is $12.33 a garment. On the reorder it is zero, which is the honest reason a custom first run costs more than a relabelled stock order and why the gap closes from the second order.

So the question is not whether custom is worth $12 a piece. It is whether you are selling your own shape at $250 retail or somebody else's shape with your label in it.

What "custom" means when a brief is not custom at all

In 2018 we quoted a custom order from a client whose reference photos were a wholesaler's stock dress shot on a different model. We drafted a new pattern to the pictures, and it fitted nothing like the photo because the photo had never been the garment.

We now ask for the measurements of the reference garment, or the body it was worn on, before we call a brief custom. Two chest, waist and hip numbers plus a length save a pattern and a mock-up round.

The test takes ten minutes on your side. Ask for the flat measurements of the reference garment and compare them with the first mock-up when it arrives. If the two do not line up, the brief was never custom, whatever the invoice said.

We decline wool outerwear, knitwear and hand beading at any volume, and our cloth list stays inside viscose, polyester, linen, cotton-linen, satin, rayon and TR suiting.

Our small batch program page sets out quantities and minimums. The FAQ page answers the custom questions we get most, and the case studies page lists delivered orders with their quantities and fit outcomes attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to separate a custom brief from a stock one before a pattern is drawn.

Sourcing calls use three words almost interchangeably — accessible luxury, premium, high-end — and buyers and factories rarely mean the same thing by any of them. When you are choosing a clothing manufacturer, the tier is not decoration: it decides which fabrics a factory stocks, which tolerances it measures, and what its minimum order value has to be.

Here is the same axis from a factory floor's point of view.

Accessible luxury: better than the price suggests

Accessible luxury — often called light luxury or contemporary — is the tier between mid-market and true luxury. Elevated basics and statement pieces, real construction, fabrics one grade above fast fashion, at retail prices a boutique shopper accepts without flinching. In sourcing terms: elevated fabrics such as cupro, silk blends and TR suiting, honest finishing, and retail roughly in the $150–$1,000 band.

For the factory, this tier lives or dies on cost discipline. The manufacturer has to know where the money shows — fabric hand, fit, finishing — and where it does not. A target-price workflow (quote backwards from the number the brand needs) is the honest way to run this tier; a factory that only quotes forward from a spec sheet will either miss the price or quietly downgrade the cloth.

Premium: the specification is the product

Premium describes the tier above mid-market where the written specification becomes the product: tolerances per size agreed before cutting, fabric weight and composition locked in the quote, stitch density on paper. A premium fashion manufacturer proves the tier with measurement rather than mood — three inspections per garment, size-by-size sheets you can audit, packing you can watch on video.

The practical test: ask what happens when a garment fails. In a premium tier factory the failure is caught by a named inspection with a written defect category, not by your customer.

High-end and luxury: the standard and the rare

High end is a production standard rather than a price tag — dynamic fit testing, tolerances that survive a pattern department's audit, finishing built for a boutique return policy. Luxury adds what is genuinely scarce: hand-guided embroidery, embellishment, silk and rare materials, and the patience they require.

One factory can run all three tiers — we grade quality in three levels and quote each honestly — but not every floor does. Many budget-line factories cannot hold a premium spec at any price, because the tolerances require the inspection system itself.

How to tell which tier a factory actually runs

Four questions sort it in one call. One: what tolerance do you hold on a shoulder point, per size? Two: what is your AQL, and is the final inspection recorded? Three: is fit checked on a mannequin or in motion — walking, sitting, squatting? Four: what is your minimum order value, and what does it cover? A factory running the upper tiers answers with numbers. A factory running one tier answers with adjectives.

Why the tier question matters more at small MOQ

At 30–50 pieces per style the setup is the same as a 500-piece run: pattern, grading, marking, line setup. The per-unit premium below 50 pieces is that fixed cost spread thin — so the tier you choose compounds. A light-luxury brand ordering 40 pieces from a budget-line floor saves on the quote and loses on the shelf: fit drifts, finishing flattens, and the retail price the brand needs becomes harder to defend.

Match the tier to the price tag you are planning, then hold the factory to that tier's measurements. That is the whole trick.

A small batch clothing manufacturer putting 50 pieces per style on a line needs about twelve weeks from deposit to delivery. That is the honest number, and it holds for a woven dress, a blouse, a linen shirt or a pair of tailored trousers.

The schedule breaks into four blocks: three weeks for tech pack review, quotation and fabric booking, three weeks for two mock-up rounds, four weeks for bulk with three inspections, and two weeks for finishing, packing and shipping. Air freight lands at about twelve weeks. Sea freight adds four to five weeks of sailing time on top.

Here is what happens inside each block, and where a 50-piece timeline actually slips.

Week 1 to 3: tech pack review and quotation for 50 pieces

We read the tech pack in two days and come back with a line-by-line quotation in two to three. The quote splits fabric, trims, cutting, sewing, finishing, inspection, packing and freight, so you can see what each line costs rather than one blended price.

Fabric booking runs in parallel, and this is the block that decides the whole calendar. Stock cloth ships in two to four days. Dyed to order starts at 300 metres per colour and adds two weeks, which is why a 50-piece run almost always stays on stock.

If your tech pack has a reference garment but no flat measurements, this week stretches. We ask for two chest, waist and hip numbers plus a length before we draft anything.

Quotation lines at 50 pieces are more sensitive to fabric than to quantity. The same dress in a 140gsm poly-crepe and a 180gsm rayon crepe can differ by $12 a garment, while doubling the run to 100 pieces moves the sewing line by about $4. The cloth call matters more than the order size.

Week 4 to 6: two mock-up rounds before any bulk cutting

Round one takes seven to ten days on the line. You get photographs on a fit model, a measurement sheet against the spec, and the garment in your hands by courier.

Fit comments come back within three days. Round two is a correction, not a redraw, and takes five to seven days. Nothing is cut for bulk until you have signed the sealed sample.

Fifty pieces is enough that a fit error repeats fifty times. That is the reason we spend three weeks here instead of one.

Week 7 to 10: bulk production with three inspections

Cutting takes three to four days for 50 pieces. A five-size nest on a 2.1-metre marker uses about 105 metres of cloth for a midi dress, and the marker is laid so the shade band stays consistent across panels.

Sewing runs eight to twelve days depending on construction. A simple viscose dress moves faster than a lined style with a bound seam.

Marker width is decided by the mill, not by us. On a 145cm roll we nest four sizes; on 150cm we fit five, and the sixth size moves to a second marker. That is a real cost difference on 50 pieces, and it is worth asking your fabric supplier for a 150cm roll before you book.

Three inspections sit inside this block: in-line at about 30% complete, end-of-line on every piece, and pre-shipment on the packed cartons. Each one produces a written report with defect counts by category.

Week 11 to 12: finishing, packing and shipping your 50 pieces

Pressing, trimming and final measurement checks take three days. Each garment is tried on a dress form and measured against the spec sheet before it is bagged.

Cartons are packed to your size ratio, photographed, and closed with a packing list inside. Air freight to the US or Europe is five to seven days plus two days of customs clearance. Sea freight is thirty to thirty-five days.

Cartons for 50 pieces are usually four to five, packed at 12 to 15 garments each depending on volume. We send carton dimensions and weights to your forwarder before the goods leave, so the booking is made on real numbers rather than an estimate.

The balance of 70% is due after you have seen the inspection report, not before.

Where a 50-piece timeline actually slips

Four things move the date, and three of them sit upstream of us. Late fit comments cost exactly the days they are late, because the line slot cannot be given away twice.

Client-supplied fabric arriving after the cutting date is the second. We ask for cloth on site seven days before cutting, which leaves room for the wash test and a shade band check.

A change of colour or fabric after round one adds two to three weeks, because the fabric booking restarts from zero. A late deposit pushes everything, since the line slot is not held until it clears.

In 2021 we quoted nine weeks for a 50-piece satin run and shipped in ten weeks and two days. The dye lot arrived four days late and we had built no float into the calendar. Every timeline we publish now carries a two-week float, stated up front.

Fifty pieces at $50 to $80 a garment in Guangzhou is a real order, not a favour. What makes it work is a calendar with float in it and a written approval at every gate.

Our small batch program page sets out quantities and minimums, and the FAQ page answers the timeline questions we get most. The case studies page lists delivered orders with their quantities and delivery dates attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to turn a launch date into a workable production calendar before a deposit is paid.

A cut and sew manufacturer cuts and sews what you send, which means the outcome of the order is decided before the first roll arrives. Nine checks cover almost every problem we have seen at that handoff.

The short answer: confirm the metres, confirm the width, wash-test for shrinkage, check the shade band, label every roll, pack for the sea, send trims with spare, agree the arrival date, and put all of it on one packing list. Miss the first two and nothing downstream can rescue the order.

Here is each check with the number attached.

What a cut and sew workshop needs before your fabric ships

Four pieces of paper. A consumption sheet with metres per garment by size, a roll list with metres and width and dye lot per roll, a wash-test result, and a packing list that matches the cartons.

Without a consumption sheet we estimate, and an estimate on a 50-piece run can be three to five metres out. That is one garment short in one size, which is the worst kind of shortage to explain to a stockist.

Shrinkage and colour: test the roll before it is cut

A wash test takes 24 hours and costs nothing at a mill that already runs them. Ask for shrinkage on the warp and the weft, and treat anything above 3% as a problem for a fitted garment.

Colour is the second half of the same test. Every roll carries a dye lot number, and two lots in one order means two shade bands. On a 50-piece run that shows up as a side seam reading darker than the bodice.

If you have to use two lots, tell us before cutting rather than after. We split them by size or by panel so the difference lands somewhere a customer will not see it.

How many metres to send for a 50-piece cut and sew order

Take the consumption, add the marker waste, then add a handling margin. A viscose midi dress uses about 2.1 metres, so 50 pieces is 105 metres before any waste.

Marker waste on a five-size nest runs 8 to 12%, which is 8 to 13 metres on that order. Send 120 metres and you have a spare garment or two in hand; send 105 and you are short on one size.

Width matters as much as length. A 145cm roll and a 150cm roll are two different markers, and a width change after the lay is planned costs a full day of cutting room time.

On a print or a one-way design, add a further 5% for matching at the seam. A repeat that has to line up across a centre-front seam can push consumption to 2.35 metres, which is 12 metres more on the same 50 pieces.

Cartons, rolls and shade bands: how to pack fabric for shipping

One roll per bag, the lot number written on the outside in marker, and rolls grouped by lot inside the carton. Never mix two lots in one bag, however close the shades look under factory light.

Fabric travelling by sea needs a moisture barrier. A roll that sits in a container for thirty days without one arrives with mildew along the fold, and mildew does not wash out of a finished garment.

Send the trims with the fabric rather than after it. Buttons, zips, interlining and thread in the same shipment, with 5% spare on anything that can break or go missing.

Nine checks before your fabric leaves your supplier

One: consumption in metres per garment by size. Two: roll count, metres and width per roll. Three: the dye lot number on every roll. Four: a wash test for shrinkage and colour fastness.

Five: weight per roll, so the freight quote is right before the pallet is built. Six: a moisture barrier on sea freight. Seven: trims shipped together with 5% spare. Eight: an arrival date agreed at least seven days before cutting.

Nine: one packing list that matches the cartons. Nine lines on a single page, and we ask for all nine on the first order rather than on the second.

What goes wrong when fabric arrives late

A late roll does not only delay the start, it loses the line slot. Our cutting room is booked two to three weeks out, and a slot given up goes straight to another order.

In 2019 a client sent six unlabelled rolls with no lot numbers on them. We cut the largest rolls first and found two of them half a shade darker halfway through the lay. Seven of forty-five skirts had to be remade from a re-ordered twelve metres, which cost nine days and started a new shade band.

We keep the roll list pinned to the cutting ticket, so when a defect turns up in a roll we know which garments carry it and how many metres are left in that lot. That traceability is what makes a remake a two-day job instead of a fresh order.

Cut and sew is a partnership with a clean split: you own the cloth, we own the making. The handoff is where the risk sits, and it is the cheapest part of the order to get right.

Our cut and sew page sets out what we do in-house and what you supply, and the FAQ page answers the fabric questions we get most. The case studies page lists delivered orders with their quantities and fabric sources attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to flag a fabric handoff that will not cut cleanly, before the rolls ever leave your supplier.

A private label clothing manufacturer quoting the same dress can land anywhere between $42 and $71, and the gap is rarely the sewing. Three quotes for one style usually differ on four lines: cloth, lining, grading and terms.

The short answer: ask each factory to break the quote into fabric, lining, trims, cutting, sewing, finishing, inspection, packing and freight. Put the three breakdowns side by side and the difference stops being a mystery and becomes a set of decisions you can actually make.

Here is where 29 dollars goes on a midi dress at 50 pieces.

Why two private label quotes are never the same garment

Same style, same 50 pieces, three quotes: $42, $58 and $71. None of the three is dishonest, and none of them is quoting the same garment.

Quote A uses a stock poly-crepe at $7.00 a metre, no lining, one mock-up and four sizes. Quote B uses a viscose at $10.00, a half lining, two mock-ups and five sizes. Quote C uses a rayon crepe at $12.85, a full cupro lining, two mock-ups, six sizes and delivers DDP.

The price difference is a specification difference. Read the specification before you read the price.

Fabric: the line that moves 40% of a private label quote

Cloth is the largest single line in a garment, and it is the one most often left vague. Two point one metres at $7.00 is $14.70; the same metres at $12.85 is $26.99. That is $12.29 on one garment before anything has been sewn.

A lining is a second fabric with a second consumption. Half lining adds about $3, a full cupro lining about $6, and both add sewing time on top of the cloth.

Weight belongs in the same conversation as price. A 140gsm crepe and a 180gsm crepe use the same metres but read as two different garments on the body, and the heavier cloth is the one that holds a pressed hem through a season of wear.

Ask for the cloth by name, its weight in grams per square metre, and its price per metre. A quote that says "polyester blend" is not a quote yet, it is a placeholder.

Labour and construction: 8 to 22 dollars of the same garment

Sewing is not one number. A single-needle seam with an overlocked edge takes minutes; a French seam, a bound seam or a hand-finished hem takes longer and reads differently on the body.

Across the three quotes, sewing runs $16, $22 and $26. That $10 spread buys cleaner insides, a straighter hem and fewer repairs after the end-of-line inspection.

Grading is the quiet line. Four sizes is four patterns; six sizes is six, plus a grade rule that has to hold at both ends of the range. More sizes means more pattern work and a more complex marker on the same 50 pieces.

Ask how many pattern pieces the style carries. A simple dress runs six to ten; a lined dress with a facing and a waist seam runs fourteen to eighteen. Every extra piece is another cut, another seam and another check at the end of the line.

Trims, labels and packaging: the small lines that add up

Quote A carries $3 of trims, quotes B and C carry $7. The difference is a woven neck label rather than a printed one, a proper care label, and a hangtag with a string rather than a plastic barb.

Packaging is where the terms hide. FOB means the garment is yours once it is on the vessel; DDP means freight, duty and clearance sit inside the number you are reading. That is $6 to $9 a garment on a 50-piece air shipment to Europe.

Neither term is better than the other. What matters is knowing which one you are comparing.

What a private label quote should include, line by line

Eight lines: fabric, lining, trims, cutting, sewing, finishing, inspection, packing. Then three below the line: mock-up rounds included, sizes graded, and the Incoterm.

Ask for the pattern and grade cost separately, and ask whether it is credited against the bulk order. A $180 pattern credited on the first order is $180 once, not $3.60 a garment forever.

Ask what the quote excludes. Wash tests, extra mock-up rounds and a second hangtag design are the three additions that most often appear after the quote has been signed.

How to compare three quotes without comparing prices

Line the three breakdowns up against the same spec sheet. Where two quotes differ by more than 15% on one line, one of them is describing a different garment rather than a better price.

In 2022 we quoted a private label run from a client's own fabric specification without re-testing the sample she sent us. The cloth shrank 6% in the wash test against our 3% limit, and the 40-piece run had to be re-cut in a substitute. We now wash-test every client-supplied fabric before the quote goes out, which adds a day and has saved more orders than it has delayed.

Three quotes are useful, but only if they describe the same garment. Ask for the breakdown, put it against your spec sheet, and the cheapest quote usually stops being the cheapest.

Our private label page sets out what we supply under your brand from 30 pieces, and the FAQ page answers the pricing questions we get most. The case studies page lists delivered orders with their quantities and target prices attached. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to read three quotes against one spec sheet and show where they stop matching.

A small batch clothing manufacturer quotes 40 pieces per style differently from 30 or 50, and the difference is arithmetic rather than goodwill. Forty is where most independent womenswear labels land for a first run, and it is the number factories handle worst.

Is 40 worth it, or should you do 30 and reorder? The short answer: 40 pays off only when you are running more than one style in the same cloth. One style at 40 pieces costs about $79 a garment from us; three styles at 40 pieces, sharing one fabric order and one cutting plan, comes down to roughly $74. One style alone is the most expensive way to buy forty garments.

Here is the arithmetic, and the three places 40 behaves differently from 30.

Why forty pieces per style sits between small batch and a real run

Thirty pieces exists so you can test a style. Three hundred exists so a factory can run one style for a week without changing a setting. Forty sits in the gap, and the gap is not a smooth slope.

A line of six people can cut and sew 40 pieces of a simple dress in four days. That line still needs the same day and a half of setting up whether the order is 30 pieces or 40, so the setup is spread over more units.

What 40 does not fix is the size range. Forty pieces across five sizes is eight garments per size, and eight is thin for a market test. Two sizes will sell out and two will sit.

If your label sells one or two shapes well, three styles at 40 pieces each is a stronger first run than 120 pieces of one dress.

Cost per piece at 30, 40 and 50 pieces from one quote

Every garment quote has a fixed block and a variable block. The variable block barely moves with quantity: 2.1 metres of viscose crepe at $8.40 is $17.64, half lining $3.20, trims $4.60, cutting and sewing $19.50, finishing, inspection and packing $6.40. That is $51.34 before anything else.

The fixed block is $1,100 for a new style: pattern $180, grading $120, two mock-up rounds $420, tech pack review $180, marker $200. Divide it by quantity and the whole cost changes shape.

At 30 pieces the fixed load is $36.67 and the garment lands at $88.01. At 40 it is $27.50 and the garment is $78.84. At 50 it is $22.00 and the garment is $73.34.

Forty is a real step down from 30 and a real step up from 50. On a three-style run, going from 30 to 40 per style adds 30 garments and about $1,540 of variable cost, while pulling $3,300 of fixed cost down to $2,750.

Fabric planning for three styles at 40 pieces each

Three styles at 40 pieces is 120 garments, and at 2.1 metres each that is 252 metres before the marker allowance. Add 8 to 12% and you need about 275 metres.

That number matters because a dyed-to-order cloth carries a 300-metre minimum at most mills. Three styles at 40 pieces gets you to the minimum without buying dead stock. One style at 40 does not.

Cutting three styles from one cloth also saves metres. Nesting a bodice, a skirt panel and a sleeve across the same lay recovers 4 to 6 metres on 120 garments.

One dye lot is one colour across the whole run. Split the order across two lots and you will see it in the finished garments, not in the rolls.

Mock-up rounds and size range at forty pieces per style

Forty pieces gets the same two mock-up rounds as 300 pieces. Round one checks the pattern and the construction, round two checks the fit on a body and the grade at both ends of the range.

Approve a range from one size and you are gambling. In 2021 a client signed off on a size 8 sample and we cut a five-size range from it. The size 14 armhole gaped by 2 centimetres, six garments had to be re-cut, and the run went out nine days late.

Now we make the round two sample in the smallest and the largest sizes of the range, not the middle one. It costs about $180 extra and it is the cheapest insurance on the quote.

What to send before we quote a forty-piece run

Four things: a tech pack or a reference garment, the size range and the split across sizes, the cloth by name and weight, and the date you need the goods in your hands.

If you have no tech pack, send a reference garment and a photograph of the detail you care about. We measure it, write the spec and send it back for approval before quoting.

Send the size split as numbers, not as a percentage. Eight, eight, eight, eight, eight is clear. "Evenly spread" is not, and it changes the marker.

What goes wrong at forty pieces and how we handle it

Cloth arrives late, a dye lot does not match the swatch, or one size sells out before the reorder window. The first two we plan for. The third we cannot.

In 2021 we cut two 40-piece styles from one lay to save the marker. The second style carried a wider skirt panel, marker efficiency fell from 89% to 81%, and we lost 11 metres and four days re-cutting. We now test the marker before the lay is confirmed rather than after.

Reorders are the real risk. A 40-piece style that sells out in three weeks needs a second run, and the second run repeats the fixed block unless the pattern is already paid for. Ask whether the pattern cost is credited against the first bulk order.

Forty pieces is a market test, not a scale play

Buy 40 pieces per style to learn which shape sells, in which size, at which price. Then put the money behind the two that worked.

Twelve weeks is a normal lead time for a 40-piece run, fourteen if the cloth is dyed to order. Thirty per cent deposit, seventy on inspection.

Our small batch programme starts at 30 pieces per style, and the FAQ page covers the quantity questions we get most. The case studies page lists delivered runs with their quantities and target prices. Our production process page shows the two mock-up rounds and three inspections every order goes through. The Fashion Pains Solver we are bringing online will read a size split and a cloth order and show which quantities waste cloth.

A high end clothing manufacturer is judged on fit before it is judged on price, and most brands approve the first sample that looks right on a hanger. Six checks take about forty minutes and catch the faults that surface at the wrong moment.

What should you actually check before you approve a womenswear sample? Measure it against the spec sheet at five points, look inside at the seams, put it on a body and wear it for an hour. A sample that passes a mirror and fails a tape measure costs a season.

Here are the six, in the order we run them at the approval meeting.

Check one: five measurements against the spec sheet

Bust, waist, hip, length and armhole. Measure the sample flat, then measure it on the body, and write both numbers down.

Our tolerance is plus or minus 0.5 centimetres on those five points and plus or minus 1 centimetre everywhere else. Symmetry is 3 millimetres, which means the left and right side seams should match each other before they match the spec.

A garment can be 1 centimetre off the sheet and still fit well. The same garment 1 centimetre off across two sizes fits badly, and the sheet is a starting point rather than a verdict.

Check two: seam allowance and the inside of the garment

Turn it inside out. Look at the seam allowance, the overlocking and the hem finish. One centimetre is our standard, and 1.5 centimetres at stress points like the armhole, the side seam at the hip and the back waist.

Stitch density belongs here too. Ten to twelve stitches per inch on a woven seam holds. Six to eight looks fine in a photograph and starts opening after three wears.

Inside finish is where a cheap garment shows itself first. Nobody sees it in a lookbook, and everybody sees it when a customer returns a dress.

Check three: dart, armhole and shoulder placement on a body

Put the garment on a fit model and look at the dart apex. On a bust dart it should sit 2 to 3 centimetres away from the bust point, not on it.

Then look at the armhole from the side. If there is a gap under the arm while the model stands still, the armhole is too low or the shoulder is too wide.

Shoulder seam placement decides whether a garment reads as tailored or as borrowed. The seam should finish at the edge of the shoulder bone, within 0.5 centimetres.

Check four: two sizes of the same style side by side

Lay the smallest and the largest size of the range on the table together. Then measure the grade step between them. Our womenswear grade runs bust plus 5 centimetres, waist plus 4.5 and hip plus 5 per size.

Grading is where womenswear fails quietly. A grade rule copied from a knit pattern onto a woven dress will hold at size 8 and collapse at size 14.

In 2021 a client approved a range from a size 8 sample and we cut five sizes from it. The size 14 armhole gaped 2 centimetres, six garments were re-cut and the run shipped nine days late. We now make the round two sample in the smallest and the largest size, which costs about $180 and has not failed since.

Check five: press the garment and watch the cloth

Press the sample on the setting you will use in production. Viscose and rayon move under heat, and a satin face can shine if the iron is too hot or the cloth is pressed damp.

Watch the hem. A hem that ripples after pressing has been cut off-grain or the cloth was not relaxed before cutting. Rippling does not settle after the first wash.

Look at the colour under two lights, ideally daylight and a shop light. Two rolls from one dye lot can still read differently under different light, which is why we check with a grey scale against a level 4 to 5.

Check six: wear it for an hour before you sign

Sit down, walk, lift your arms, climb stairs. An hour on a body finds the pulling at the back waist and the hem that rides up, and neither shows up on a mannequin.

Every garment in a bulk run is tried on a mannequin at the end of the line, and one in ten is tried on a fit model. That is our floor, not the standard we aim for.

If a garment fails at this stage it is still a sample and it costs a week. If it fails after 300 pieces are packed, it costs the season.

What these six checks cannot tell you

They cannot tell you whether the cloth will still look right after twenty washes. A wash test at the start of the run answers that, and it has to happen before cutting.

They cannot tell you whether the customer will like it. Fit is measurable, taste is not, and no factory can grade taste.

What they can do is remove the four faults that come back most often: a grade that collapses at the top of the range, a seam that opens, a hem that ripples and a dart that sits in the wrong place.

Our high end production page sets out the standards we hold, and the FAQ page answers the fit questions we get most. The case studies page lists delivered runs with the fit problems we solved on each. Our production process page maps the two mock-up rounds and the three inspections. The Fashion Pains Solver we are bringing online is being built to read a sample measurement sheet against a spec sheet and flag the points outside tolerance.

A private label clothing manufacturer quotes FOB unless you ask for something else, and FOB ends at the port. The number on the quote is not the number you pay, and the gap is wider on small orders than on large ones.

Which term should a first order use? FOB if you already have a freight forwarder and a customs broker, DDP if you want one number and no paperwork. On 120 pieces the difference runs 8 to 14% of the garment cost, and DDP is not automatically the expensive choice once you price your own hours into it.

Here is what sits inside each number, and where small brands lose money without noticing.

What FOB and DDP actually include on a garment order

FOB means the goods are yours once they are loaded at the Chinese port named on the quote. Freight, insurance, duty, clearance and delivery are yours to arrange and pay.

DDP means we carry the goods to your door, cleared and duty paid. One number, one invoice, and the customs paperwork is ours.

EXW is the third term you will see and the one to avoid on a first order. EXW means the goods are yours at our door in Guangzhou, which leaves you arranging export clearance from another country.

Freight at 42 kilograms: air against sea for small brands

One hundred and twenty garments at 0.35 kilograms each is 42 kilograms of product, and about 48 kilograms once cartons and volumetric weight are counted.

Air freight runs $6.50 to $9.00 a kilogram to Europe or the United States, so $312 to $432, and lands in five to seven days. Sea LCL is $180 to $260 plus $120 to $180 of local charges at both ends, and takes five to eight weeks door to door.

Below 100 kilograms the local charges at both ends eat the saving on sea. Above 300 kilograms the difference becomes real money.

Air freight is where a launch date gets rescued. Sea is where a repeat order gets its margin back.

Duty and VAT: who pays and when it comes back

Duty is charged on the customs value, usually the FOB value plus freight and insurance. The United States sits around 16% on most apparel lines, the European Union around 12% and the United Kingdom around 12%.

Those rates move, and they depend on the HS code, which depends on the garment. A viscose dress and a polyester dress are not always the same line. Check the code for your specific garment before you build a retail price.

VAT is a different thing from duty, and this is the part small brands miss. UK VAT is 20%, Italy 22%, France 20% and Germany 19%, and a VAT-registered business reclaims it. Duty never comes back.

If you are not registered for VAT the tax is a real cost. If you are, it is a cash-flow gap of a few weeks rather than a cost.

Customs value: the line nobody puts on the quote

Customs value is not the number you feel like declaring. It is the transaction value, and it should match the invoice and the payment.

Under-declaring saves duty on paper and creates the risk of a hold, a fine or a revaluation. A container held for three weeks costs more than the duty it saved.

Ask for the commercial invoice, the packing list and the HS code before the goods ship, not after. DDP does not remove the need to know your own numbers.

Worked example: 120 pieces, FOB against DDP landed

Three styles at 40 pieces each, $78.84 FOB a garment, which is $9,460.80. Air freight and insurance add $380. Duty at 12% on $9,840.80 is $1,180.90. Clearance is $150 and local delivery $90.

Landed cost is $11,261.70, or $93.85 a garment. A DDP quote on the same run at $101.00 a garment is $12,120, which is $858 more, or $7.15 a garment.

That $7.15 buys the customs entry, the duty payment, the clearance agent and the delivery. If your own time on those four items is worth more than $858, DDP is the cheaper term.

In 2020 we quoted FOB to a first-time importer who read the number as a delivered price. Her shipment sat at the port for nine days while she found a broker, and the storage bill was $340. We now write the Incoterm in capitals at the top of every quote and repeat it in the covering email.

Which term fits a first order and which fits a repeat

First order, DDP. You are buying certainty and you have no forwarder relationship yet. The premium is small next to the cost of a delayed launch.

Repeat orders, FOB. By then you have a broker, you know your HS code, and you can consolidate two or three shipments into one customs entry.

If you ship into a fulfilment centre, check what the centre will accept. Some require DDP delivery and will refuse a shipment that arrives with duty owing.

Three questions to ask before you accept either term

Which port or airport, exactly. What the quote excludes. And what happens if the goods are held.

A quote that names a port, lists its exclusions and states who pays storage on a hold is a quote you can compare against another one. A quote that says "shipping included" is not.

Our private label page sets out what we supply under your brand from 30 pieces, and the FAQ page covers the terms questions we get most. The case studies page lists delivered orders with their Incoterms and clearance times. Our production process page maps the two mock-up rounds and three inspections that sit before any shipment. The Fashion Pains Solver we are bringing online will read two quotes against each other and show which lines sit inside the price and which do not.

A small batch clothing manufacturer charges less for your second run than your first, and the reason is arithmetic rather than goodwill. The pattern, the grading and the marker are already paid for.

How much less, and how much faster? On a 30-piece viscose dress we quote about $88 the first time and about $72 the second, and the lead time drops from twelve weeks to five. That is an 18% drop on the garment and seven weeks back in your calendar.

Here is what carries over, what does not, and the one mistake that turns a reorder into a fresh first order.

What a repeat order does not pay for the second time

Every new style carries a fixed block before a single garment is sewn. On our quotes that block is $1,100: pattern $180, grading $120, two mock-up rounds $420, tech pack review $180, marker $200.

None of it repeats if the style is unchanged. The pattern is in our system, the grade is approved, the sample is signed and the marker file is saved.

Spread over 30 pieces, that $1,100 was $36.67 a garment. On a reorder it is zero, and that alone is most of the gap between $88 and $72.

The rest of the gap is the sample rounds. A reorder in the same fabric and the same colour goes straight to bulk cutting.

Second run lead time: five weeks against twelve

A first run takes twelve weeks because the mock-up rounds sit in the middle of it. Two rounds with shipping both ways is four to five weeks on its own.

A reorder skips them. Three weeks of fabric and trims, four weeks of cutting and sewing, one week of finishing, inspection and packing.

Five weeks is realistic when the fabric is in stock. If the colour has to be dyed again, add two weeks and the advantage shrinks.

So the planning numbers are: reorder in a stock colour, five weeks. Reorder in a dyed-to-order colour, seven to eight.

The four records that decide whether a reorder is cheap

A reorder is cheap only if the records survived the first run. Four things have to be kept.

The approved sample, not a photograph of it. The signed spec sheet with the tolerances written on it. The fabric card with the mill, the article number and the dye lot. The trims card with the supplier reference for every zip, button and label.

Without the fabric card we cannot match the colour. A mill discontinues a shade roughly once every two years, and matching from a finished garment is guesswork.

In 2019 a client reordered a viscose dress eight months after the first run and asked for "the same blue". We had no fabric card on file. The new dye lot came out a shade deeper, and she had 40 dresses that did not sit next to the 30 still on her rail.

Every order now leaves with a one-page record sheet. It takes twenty minutes to build and it is the difference between a five-week reorder and a twelve-week rebuild.

How to read sell-through before you commit to a reorder

Reorder when a size sells out, not when a style feels popular. Sell-through at 60% with two sizes gone is a signal. Sell-through at 60% with every size still on the rail at half depth is not.

Count by size, not by style. If the medium and the large went first, the reorder is not a copy of the first run.

On 30 pieces across five sizes you had six a size. A reorder can shift that to four of the slow sizes and nine of the fast ones without changing the total quantity or the price.

Rebalancing the size split is the cheapest thing a reorder does. It costs nothing and it moves the sell-through.

Reorder quantity: why 30 still beats 50 on a second run

Our minimum does not change on a reorder: 30 pieces a style. What changes is what those 30 pieces cost.

The variable cost per garment barely moves with quantity. Viscose crepe 2.1 metres at $8.40 is $17.64, half lining $3.20, trims $4.60, cutting and sewing $19.50, finishing and packing $6.40. That is $51.34.

On a first order the fixed block pushed that to $88.01. On a reorder it stays at $51.34 plus margin, so a 50-piece reorder only saves the fabric difference from a tighter marker.

Going from 30 to 50 on a reorder is worth about $1.10 a garment. On a first order the same jump was worth $14.67. That gap is the whole point of the second run.

When a reorder is the wrong decision

A reorder is wrong when the style sold because of a moment. A print tied to one season, a colour that was everywhere for eight weeks, a shape three other labels launched at the same time.

It is also wrong when the first run exposed a fit problem you have not fixed. Reordering a garment with a known fault multiplies the fault across two batches.

And it is wrong when the fabric is gone and the substitute is close rather than the same. Close is how two dye lots end up on one rail.

Ask for fabric availability before you commit. We check stock before we quote a reorder, not after.

Our small batch programme keeps the pattern, the marker and the fabric card on file for three years, and the FAQ page answers the reorder questions we get most. The case studies page lists the repeat orders we have shipped with their quantities and target prices. Our production process page shows which of the three stages a reorder skips. The Fashion Pains Solver we are bringing online will read sell-through by size and suggest the split for the next run.

A cut and sew manufacturer in Guangzhou runs on a calendar that has nothing to do with your launch date, and August is the month where that hurts most.

Why is August bad? Every brand that wants stock on a shelf for northern autumn places its order in the same six weeks, and the sample rooms fill before the sewing lines do. An order placed in August comes back two to four weeks later than the identical order placed in May.

Here is the year as a factory actually sees it, and how to count backwards from a shelf date.

Chinese New Year factory closure: three weeks the calendar stops

Chinese New Year is the single biggest block in the year. Factories close for two to three weeks, and the dates move with the lunar calendar, landing anywhere from late January to late February.

The week after reopening is not a full week. Staff return in stages, and a line that normally runs 40 machines often has 28 running on day one.

Mills close earlier than factories and reopen later. Fabric ordered in the last week before the holiday will not be woven until the second week after it.

If you need stock in March, the fabric has to be on our floor by mid-January. That is the real deadline, not the holiday itself.

Golden Week and the October capacity crunch

The first week of October is a national holiday, and the week before it is the busiest of the year.

Brands that need stock for the November and December selling season all want to ship in September. Sample rooms are booked three weeks out and air freight rates rise 20 to 40%.

Golden Week itself is shorter than Chinese New Year, usually seven days plus travel either side. The capacity crunch around it lasts a month.

If your launch is spring, none of this touches you. If it is autumn, every decision you make in June and July is really an October decision.

August and the sample room: why samples queue before the line does

August is not a holiday month. It is the month before the rush, and that is why it is the worst time to start.

Sample rooms run at the front of every order. A mock-up round takes ten to fourteen days in a quiet month and three to four weeks in August, because every autumn order wants its sample approved in the same window.

The sewing line still has capacity. The sample room, the pattern maker and the fit technician do not, and those three are what a new style needs first.

The honest version we give clients: place the order in August if your shelf date is February. If it is October, you are already late and air freight is the only lever left.

Typhoon season in Guangdong and the summer buffer

Guangdong's typhoon season runs June to September. Most years it means two or three days of lost production, not a week.

The cost is not the lost days. It is the fabric. A container delayed by a port closure arrives wet or arrives late, and both cost more than the days.

We ask for sea shipments to carry moisture barrier film year-round, and we schedule deliveries with a seven-day buffer between June and September.

Small brands rarely lose a whole order to weather. They lose four to six days, and four to six days is enough to miss a buying appointment.

Working backwards from a shelf date to an order date

Take your shelf date and count back. Twelve weeks for a first run, five for a reorder, plus two if the fabric is dyed to order.

Then add the buffer: two weeks if the order lands between June and September, three weeks if it crosses Chinese New Year or Golden Week.

A spring drop that needs stock on 1 March in a dyed-to-order fabric: 1 March minus twelve weeks is 8 December, minus two for dyeing is 24 November, minus three for the holiday is 3 November. That is your order date.

Write it in the calendar as the order date, not the ship date. Ship dates are the number that makes brands late.

What to do when you are already late for the season

Three levers, in order of how much they hurt. Split the order and ship 40% by air for the launch. Simplify and drop the print or the custom trim that adds a supplier. Extend the drop two weeks and use the gap for pre-orders.

What does not work is asking a factory to compress. A twelve-week run is not a nine-week run with more effort. The mock-up rounds sit on the calendar, not on the effort.

In 2022 we accepted an August order with an October shelf date and told the client it would ship in the last week of October. It shipped on 26 October, the launch moved eleven days, and she spent the gap on a pre-order list that sold 40% of the run before it landed.

That is the honest version of a late order: not rescued, but recoverable.

Our cut and sew page sets out the four workstations and what you supply, and the FAQ page covers the timeline questions we get most. The case studies page lists delivered runs with their actual ship dates. Our production process page maps the two mock-up rounds and three inspections that sit between a deposit and a shipment. The Fashion Pains Solver we are bringing online will read a shelf date and count back through the holidays for you.

A private label clothing manufacturer will sew your label into a garment, but the size chart on your website is a decision only you can make. Most first-time brands get it wrong by copying ours.

Whose chart should you sell? Yours, built from the block you approved, with the tolerances you can actually hold. A chart copied from a factory's base block describes the factory, not your brand, and it breaks the moment you change a fabric or a fit.

Here is what a size chart is made of, what to publish, and the three ways a chart goes wrong in a first season.

Base size, grade and the chart that comes out of them

Every garment starts from a base size. Ours is a size 8 on an EU 38 block, fitted on a model at 86 / 68 / 94 and 168 centimetres.

The base size is measured and approved. Every other size is generated by the grade, which is the increment added at each point of measure as you step up or down.

A womenswear grade adds about 5 centimetres at the bust, 4.5 at the waist and 5 at the hip per size. Those numbers are not universal. They are a house decision, and two factories can grade the same base size into two different size 14s.

Your chart is the output of that decision. Publish it as finished garment measurements, not body measurements, because finished measurements are what you can check on a garment you are holding.

EU, US, UK and Italian size labels: what goes on the tag

Size labels are not a translation exercise. A size 8 in the US is not a size 8 in the UK, and neither is an EU 38.

Working conversions for womenswear: US 8 is a UK 12, an EU 38 and an Italian 42. US 4 is a UK 8, an EU 36 and an Italian 40.

If you sell in more than one market, print both numbers on the care and content label or on a separate size label. EU labelling rules want the size in the format that market recognises, and a chart on your website does not replace the tag.

We print the tag you specify. We do not pick your size system for you, and a factory that offers to "just use standard sizes" is offering to make that decision for you.

What to publish: garment measurement, tolerance and how to measure

Publish three columns per size: the point of measure, the garment measurement and the tolerance.

Our tolerances are plus or minus 0.5 centimetres on the five key points, which are bust, waist, hip, shoulder and body length, and plus or minus 1 centimetre on everything else. Left to right symmetry is 3 millimetres.

Publishing the tolerance protects you. A customer who measures a 96 centimetre bust on a garment labelled 95.5 is inside the spec, and without the tolerance on your page you will process that return.

Say how you measure, too. Bust measured flat and doubled is a different number from bust measured around a body, and on a 40-piece run that difference becomes a stream of emails.

Vanity sizing and what it costs a small brand

Vanity sizing means labelling a garment one or two sizes smaller than it measures, so a customer feels better about the number on the tag.

It works in a large catalogue with a returns department. It does not work on a small run.

Your chart and your grade are the same decision. Shift the label without shifting the pattern and size 10 and size 12 become the same garment, which halves your range. Shift the pattern and you need a re-grade, which is a new block rather than a new label.

We quote what the block measures. If you want a vanity chart, we will tell you what it costs in grading before we accept it.

Thirty pieces across five sizes: what the chart does to the split

Thirty pieces across five sizes is six garments a size. That is thin, and the chart is what decides whether the split is worth it.

If your chart sells three sizes and not two, running five sizes at 30 pieces means carrying the two dead sizes through the whole run.

A cleaner first order for a new label is three sizes at ten pieces. You get ten of each of the sizes your chart actually sells, and you learn more from the reorder than from a full range.

Rebalancing later is free. The size split on a reorder is a spreadsheet decision, not a change to the pattern.

Why the chart has to match the fabric, not just the body

A size chart is written once and applied to a block, but the garment is made in a fabric that moves.

Viscose crepe relaxes. Linen shrinks 2 to 3% on the first wash unless it is pre-shrunk. A satin bias panel drops two days after it is cut. The same approved chart can fit differently in two fabrics.

Our answer is to write the chart on the finished garment, after the fabric is washed and the garment has rested. On a first order in a new fabric we measure the approved sample after 24 hours and adjust the chart before bulk.

In 2021 we published a chart from a mock-up measured straight off the table. The linen relaxed over the next two days and 12 of 40 dresses measured short against the chart. We re-measured after resting, reprinted the chart and changed the process. The chart now goes out only after the sample has rested a full day.

Our private label page sets out what we supply under your brand from 30 pieces, and the FAQ page answers the sizing questions we get most. The case studies page lists delivered runs with the fit problems we solved on each. Our production process page shows where the fit is checked across the two mock-up rounds. The Fashion Pains Solver we are bringing online will read a size chart against a spec sheet and flag the points outside tolerance.

A small batch clothing manufacturer charges for samples before there is any bulk order, and at 30 pieces that bill arrives long before a garment is sold. Our fees run from $80 for an unlined top to $260 for a lined dress, and we deduct them from the bulk invoice once you order.

So what will yours cost? Between $150 and $520 per style before bulk, covering two mock-up rounds and the sealed sample you approve. Three things move that number: whether the pattern is new, how much sample yardage the style eats, and how many revisions you ask for after round one.

Here is what the fee covers, when it comes back, and when we charge a third time.

What the sample fee actually covers

Four things sit inside a sample fee: pattern work, sample yardage, sample room hours, and trims.

Pattern work is the biggest variable. Drafting a new style from a tech pack is $120 to $260, because it means a new block, a marker and a grade set. Adapting a block we already hold is $40 to $90. If you bring a finished pattern from another factory, we check it rather than redraw it, usually for $40.

Sample yardage is billed at sample price, not bulk price. A mill will not open a roll for 4 metres, so you pay the agent's cut, roughly 15 to 30% over the bulk rate. A bias-cut satin dress can eat 4 metres in sampling, which is $60 to $90 before anyone sews.

The sample room is the fairest line to argue about. One mock-up on a lined dress is two to three hours of a senior machinist plus an hour of fitting. We do not mark that up, and we will show you the hours if you ask.

Two mock-up rounds at 30 pieces: our real numbers

Here is a run we quote often. A viscose crepe midi dress, 30 pieces, one colourway.

Round one mock-up is $140. Round two is $120, because the pattern is already cut and only the corrections are new. The sealed sample is $260, which includes the final fabric, the full construction and the trims. Total before bulk: $520.

At $81 a garment the order is $2,430, so sampling is $17.33 per piece of cash spent before a single dress exists. Add the 30% deposit and roughly $1,250 leaves your account before we cut anything.

We put that number in the first call, not in week five. Brands that are surprised by sampling costs are usually brands that were quoted a unit price and nothing else.

Is the sample fee refundable when you place the bulk order?

With us, yes, and it is worth asking any factory the same question in writing.

Our terms are simple. Sample fees are deducted from the bulk invoice if the order is placed within 60 days of sample approval. If you walk away after approval, you keep the sample and we keep the fee, because the work is done either way.

Some factories refund, some do not, and a few advertise a free sample that turns out to be a stock garment in their own fabric rather than your design. A stock sample tells you nothing about your fit, your cloth or your grade.

Ask one question: does the fee come off the bulk invoice, and inside what window? If the answer is vague, treat the fee as sunk.

Why a free sample usually is not free

A factory offering a free sample has to recover it somewhere. It turns up in three places.

In the unit price. Ask for the quote with and without the free sample and compare the two lines.

In the fabric. A free sample in a stock cloth tests construction and nothing else. Your fabric may shrink, drape and crease differently, so the sample you approved is not the garment you receive.

In the volume. Free sampling is easy to justify at 500 pieces and hard at 30, which is why the offer often disappears once you confirm your quantity.

We charge, and we say why. It is cheaper for both of us than a quote with a hidden line in it.

When we charge for a third round

A third mock-up means something was not decided at round one, and we quote it separately at $80 to $160.

The usual causes are a fabric swap after round one, a fit change on a size that was never fitted, or feedback from three people with three different opinions. The last one is the expensive one. We ask for one decision-maker and one consolidated comment sheet per round.

We also charge when comments arrive after the sample has been cut for round two. That is a re-cut rather than a revision, and it is priced as one.

In 2020 we quoted a bias-cut satin dress at $90 for sampling and lost money on it. The sample yardage alone was $62, and the client asked for two revisions we had not priced. The fee schedule we use now has the revision rounds written into it.

What we agree in writing before sampling starts

Four lines on one page, before any fabric is cut.

The fee per round and per sample, and what each one includes.

The refund window, and how the deduction appears on the bulk invoice.

How many revision rounds are covered, and the price of a round beyond that.

Who owns the pattern and the grade set once the sampling is paid for. That last line matters more than the fee, and we have written a separate guide on it.

Our private label page sets out what we produce under your brand from 30 pieces. The FAQ page covers deposits, payment and sampling questions, and the case studies page lists delivered runs. Our production process page shows where the two mock-up rounds sit in the calendar. The Fashion Pains Solver we are bringing online will read a sample comment sheet against a spec and flag the points nobody answered.

Letting a cut and sew manufacturer buy your cloth usually costs $0.60 to $1.40 more per metre than buying it yourself, and on a 40-piece run that is $60 to $140. We still recommend it for most first orders, because the metre you save is the metre that carries the risk.

So which should you do? Buy it yourself when you already own the cloth, when a mill has a fabric we cannot buy at our volume, or when you have found a real deadstock lot. Otherwise let us buy it, and spend your week on the design instead of on freight quotes.

Here is what actually changes when the cloth arrives from you rather than from our mill.

What changes when you supply your own fabric

Three things move to your side of the table, and none of them show up on the mill's invoice.

The shortage risk moves. We cut what is on the table, so if the roll is 6 metres short, the last size does not get made. That conversation happens in week nine rather than week one, and there is no good answer at that point.

The shrinkage risk moves. Unless the roll has been tested, we cut on the assumption that it behaves like the last batch. A 4% shrink on a 68 centimetre skirt length is 2.7 centimetres, which is more than five times our tolerance.

The arrival date moves. Fabric that lands after the cut date pushes the whole run, and by then the sample room and the cutting table have been booked for your order.

Four costs self-supply moves onto your side of the table

Fabric is 30 to 45% of a garment's FOB price, so it looks like the obvious place to save. The saving is real, and so are four costs that never appear on the mill's invoice.

Freight. 100 metres of viscose weighs around 20 kilograms. Air courier from a Chinese mill to the US is $90 to $140, and sea freight is cheaper but adds three to four weeks you may not have.

Your time. Chasing a mill, checking a roll ticket, booking a courier and answering our questions is a day of your week, and it lands on the week you least want it.

Storage. If you buy 200 metres to reach a mill minimum and use 92, the other 108 metres live somewhere. Most new brands store them in an apartment.

Emergency buys. When a roll comes up short, the replacement is bought at retail, in a hurry, with air freight attached. Six metres at $6.80 plus $45 of courier is $86, which is more than the saving on the whole order.

When letting the factory buy the cloth is cheaper

We buy at our own volume, so the gap between our price and a mill's price to you at 100 metres is smaller than it looks. On viscose crepe it is around $0.90 a metre.

For that $0.90 you get a wash test on every incoming roll, a fabric card with lot numbers kept on file, the shortage risk on our side, and one invoice instead of three.

If a roll fails the shrink test we stop and tell you before cutting, which is the entire point. Finding out after 40 garments are sewn means finding out about a re-cut.

The $0.90 also buys the thing that actually costs money later: a reorder that matches. Cloth bought in two batches from two sources is the most common reason a second run does not match the first.

A 40-piece worked example, both ways

One viscose crepe dress, 40 pieces, one colourway. Consumption is 2.1 metres a unit, so 84 metres, plus 8 to 12% for the marker, so 92 metres. Round to 100 metres for the mill minimum.

Self-supply: 100 metres at $4.20 is $420, plus $90 of air freight, so $510, with the shortage risk on your side.

Factory supply: 100 metres at $5.10 is $510, all in, with the wash test, the lot numbers and the shortage risk on ours.

The two numbers are identical, and that is not a trick to make our option look better. The mill's direct price is lower per metre, and the freight closes the gap at this volume.

Self-supply starts winning at 300 metres or more, where freight spreads thin, or on a genuine deadstock lot at a real discount. Both are good reasons. "It looked cheaper per metre" is not one.

What we need from you either way

Fabric that arrives without paperwork is the most common cause of a mismatched run, so we ask for five things when you supply.

A consumption figure per unit, roll tickets with lot numbers, composition and width in centimetres, a wash test result, and the cloth in our hands seven days before the cut date.

If you cannot get a wash test, say so. We will run one, it takes a day, and it costs nothing next to a re-cut.

In 2022 a client supplied 90 metres of linen for a 40-piece run with a mill report attached. We cut without re-testing because the paperwork looked complete. The report covered a different batch, the linen shrank 4.1%, and nine of the 40 dresses measured short against the size chart. Every incoming roll now gets its own wash test, whatever the paperwork says.

Our cut and sew page covers what we cut, sew and finish from 30 pieces. The FAQ page answers fabric and minimum questions, and the case studies page lists runs with the fabric problems we caught. Our production process page shows where incoming cloth is inspected. The Fashion Pains Solver we are bringing online will check a roll ticket against the spec sheet before the roll is cut.

A private label clothing manufacturer will develop a pattern for your style and keep it on file, but who owns that pattern is a different question from who paid for it. Our answer is plain: if you paid the development fee, the pattern is yours on request.

So who owns what? You own the tech pack, the label design, the artwork and any pattern you paid to develop. We own the base block the pattern was drafted from and the CAD library it lives in. Most disputes start on the line between those two.

Here is where that line sits, and what actually travels if you ever move a style.

Who owns the pattern when you pay for it

Three files get confused with each other, and the confusion is expensive.

The pattern is the set of pieces cut for your style in your size run. You paid $120 to $260 for a new one, or $40 to $90 to adapt a block we already hold.

The block is the base size everything else is graded from. Ours is a size 8 on an EU 38 block, 86 / 68 / 94 at 168 centimetres, built up over 20 years of fit decisions.

The grade set is the output that turns one approved size into five.

Our position: the pattern and the grade set for your style are yours. The block stays ours. We will license the output of it, and we will not hand over 20 years of accumulated fit because one style was developed on top of it.

If you want the pattern, ask for it in writing when you approve the sample. We export it at approval now, rather than waiting for someone to request it.

The tech pack, the marker and the block: three different files

Your tech pack is a document. You wrote it or paid someone to write it, so it is yours, and we treat it as confidential. We do not quote your style to another brand, and we will put that in writing.

A marker is a layout for one order. It depends on the fabric width and the size split you actually ordered, so it is not portable. Sending a marker to another factory tells them almost nothing useful.

The block is the fit history. It is the reason a size 14 still fits when a size 8 was the one approved. It is also the one thing we will not sell, because it is the workshop.

Your label, your artwork and the files we keep

Your logo, label artwork, care and content text and hangtag design stay yours. We print them, we do not own them.

Two practical points. We keep your artwork on file for repeat orders and will delete it on written request. We also hold the trim supplier's artwork, which matters when you reorder a year later and the file is the only thing standing between a matching label and a 500-piece minimum.

Care and content text is a legal document in most markets rather than a design file. We print what you specify and flag it when something looks missing, but the compliance call is yours to make.

Can you move your pattern to another factory?

Yes, and we will help. It takes two to four weeks, and what it costs you is the fit history.

What travels: the pattern file, the graded size set, the spec sheet with tolerances, and the sealed sample. Without the graded set the new factory re-grades from your base size and the larger sizes move. Without the spec sheet they re-measure the sample and guess at the tolerance.

What does not travel: the fit notes from two mock-up rounds, the fabric card, and the reason three corrections were made in round two. A new factory starts from the sample, so any problem the sample hides comes back with it.

In 2021 a client moved a style to a factory in Portugal. We took 11 days to send the pattern because the file was in an old CAD format and had to be re-exported. Eleven days was our fault, and it is why the file now goes out at approval rather than on request.

Why the ownership question is worth money at 30 pieces

On a 30-piece run at $80 a garment the order is $2,400. A new pattern at $180 and a grade set at $120 is $300, which is 12.5% of the value of the order.

That is why the question is not academic at this size. You are paying for an asset, and an asset should sit on your side of the ledger.

The second order is where it pays off. The pattern is already yours and already graded, so the $300 is gone from the quote and the unit price drops.

What we put in writing about ownership

Four lines, in the sample approval email, before bulk.

The pattern and grade set for your style belong to you, and we export them at approval.

The base block and the CAD library stay with the workshop.

Your tech pack, artwork and label design are confidential and will not be shown to another client.

Your style is not produced for anyone else while you are ordering it.

What we will not sign is blanket exclusivity over a fabric or a category. A mill sells its cloth to everyone, and promising otherwise would be a promise we could not keep.

Our private label page sets out what we produce under your brand from 30 pieces. The FAQ page answers ownership and confidentiality questions, and the case studies page lists delivered runs. Our production process page shows where the pattern and the grade set are approved. The Fashion Pains Solver we are bringing online will read a tech pack against our pattern files and flag the points of measure with no tolerance attached.

A small batch clothing manufacturer quotes a minimum order quantity as a starting position, not a rule written into the building. Ours is 30 pieces per style, and we have gone below it when the levers below made sense. So when a factory says 300, ask which part of that number is the machine and which part is the margin.

The short answer is that most minimums are set by setup cost, not by capacity. A line that has to change over twelve stations, buy 1,000 woven labels and dye 300 metres of cloth is expensive to start. Remove any one of those three and the number drops. Here are the five levers we use, in the order they save the most.

What actually sets a manufacturer's minimum order quantity

Two blocks sit behind every quote. Fixed cost, which does not scale with quantity, and variable cost, which does.

Fixed cost on a lined style runs about $1,100 before a single garment exists. Pattern $180, grade set $120, two mock-up rounds $420, tech pack work $180, marker $200.

Divide that by 30 and it is $36.67 a piece. Divide it by 300 and it is $3.67. The work is identical either way, and that one division is most of the difference between the two quotes in your inbox.

So the useful question is never whether we will lower our minimum. It is which fixed cost can be removed, shared or delayed.

Lever one: stock fabric instead of dyed-to-order cloth

This is the biggest lever and the least popular, because it means changing your colour.

Dyed-to-order has a mill minimum of about 300 metres. On a 50-piece dress at 2.1 metres a piece you need 105 metres, plus 8 to 12 percent for marker waste. You are buying three times what you need, and the leftover sits on a shelf in Guangzhou.

Stock fabric has no minimum. Buy 120 metres of a stocked viscose and the gate disappears. The trade is real: you pick from the mill's colour card instead of your own.

Lever two: fewer colours, more sizes per colour

Every colour is its own small order. Two colours at 30 pieces each is two setups, two markers and two cutting lays.

One colour at 60 pieces in five sizes costs close to a single 30-piece run. If you need a second colour, run it as a reorder once the first sells. We have watched brands spend a whole launch budget on a colour that did not sell.

Five sizes at 30 pieces is six garments a size, which is thin for a first read. Three sizes at ten is often the better split.

Lever three: batch two styles on one marker

A marker is the layout that decides how much cloth you waste. Nesting five sizes of one style recovers 4 to 5 metres against cutting them separately, roughly $18 to $25 on a 50-piece run.

Batch two styles that share a fabric and a colour and the recovery is larger. You lose the ability to pull one style mid-run, so only do this once both are approved.

Lever four: accept a longer window

A small line gets scheduled into gaps between larger orders. That means a date we pick, not a date you pick.

Give us a three-week window instead of a fixed Friday and we can often fit 30 pieces where 300 would not go. What you cannot ask for is the small-run price and the large-run date.

One more note on windows. A date that is two weeks out is a favour, and a date that is eight weeks out is a plan. Ask for the second and you will usually get it.

Lever five: pay the fixed cost once, in writing

Development is the expensive part of a first order and the cheapest part of a second. The pattern and the grade set carry over. The two mock-up rounds do not, but the corrections they produced do.

Our second-run quote on the same style comes in about 18 percent below the first, and five weeks instead of twelve. That is not a discount. It is work that no longer has to happen.

When a 300-piece minimum is the honest answer

Three cases. Custom dyed cloth in a colour only you want. Custom hardware, where the mould starts at 1,000 pieces. And custom jacquard, where the loom setup begins at 800 metres.

In those cases we say so on the first call rather than take the order and quietly substitute something cheaper. In 2019 we did the second thing for a client on a custom-dyed satin. The substitute arrived two shades off under daylight and 40 garments were re-cut. We have not repeated it.

What we will not do to reach your number

We will not skip the second mock-up round, drop an inspection, or shave the seam allowance. Those are the three things that make a cheap order expensive later.

What we will do is put the fixed cost on its own line so you can see what you are buying. A quote that hides setup inside the unit price is a quote you cannot negotiate.

Our small batch program sets out what we run from 30 pieces. The FAQ page answers minimum order questions directly, and the case studies page lists delivered runs with their real quantities. Our production process page shows where each fixed cost is incurred. The Fashion Pains Solver we are bringing online will read a tech pack against our cost model and show which levers apply to your style before you have to ask.

A cut and sew manufacturer puts a lead time on the quote, and that number assumes everything arrives on time. Ours is twelve weeks for a first run from deposit to shipping, and eight to ten for a reorder. When a run slides to sixteen, it is almost never the sewing that did it.

The honest breakdown is four blocks, and only one of them is time on a machine. Three weeks for quotation, fabric booking and lab dips. Three weeks for two mock-up rounds. Four weeks on the line, and two for finishing, packing and the shipping window.

Almost everything that goes wrong happens in the first six weeks rather than the last four. That is the part worth planning around.

How twelve weeks actually breaks down

Three of the twelve weeks pass before anyone touches a sewing machine. That surprises brands who expect the clock to start at the cutting table.

Week one starts the day the deposit clears and the fabric order is placed, not the day you approve the sample. Brands who count from sample approval lose three weeks on paper before anything has actually gone wrong.

Quotation and fabric booking is two to five days on our side, then the mill's own lead time on yours. Stock cloth ships in three to five days. Dyed-to-order adds two weeks for dyeing and one for lab dips you have to approve.

If you approve the lab dip late, the whole schedule moves by that delay. It is the single most common cause of a late first run.

Two mock-up rounds, and what a third one costs

Two rounds is our standard, and it is where fit gets fixed. Round one is the first fit on a live model. Round two confirms the corrections and becomes the sealed sample.

A third round adds ten to fourteen days, because the pattern has to be corrected, the fabric re-cut and the sample room re-queued. It happens when the comments on round two are contradictory rather than specific.

"Make it tighter but keep the ease" is not a comment we can act on. "Take 1.5 centimetres out of the waist at the side seam" is.

Line scheduling: where a small run waits

This is the block brands underestimate most. A 30-piece run does not get its own line for four weeks. It gets scheduled into gaps between larger orders, and those gaps are set about three weeks ahead.

So the date you get is the date the line can take you, not the date you asked for. In August, and again in the three weeks before Chinese New Year, the gaps close and small runs wait longer.

Two practical notes. A deposit does not buy a slot until the fabric is confirmed, because a line with no cloth is a line that gets bumped. And a public holiday week is not a working week, whatever the calendar says.

We will tell you the window before you pay a deposit. If your launch date cannot absorb it, better to know then than in week eleven.

Inspection, rework and the week nobody plans

Three inspections sit inside the four-week line block, not after it. Inline during sewing, end of line on the finished garment, and pre-shipment on packed cartons.

Rework is the wildcard. Most runs need two to four days of it. A run that fails the first inline check needs a week, and that week comes out of the schedule rather than out of thin air.

We will not ship a run that fails the inline check without telling you. On 30 pieces it usually means one machine's tension was off rather than the whole run being wrong. We fix the machine, re-cut the affected pieces, and the schedule absorbs the days.

Shipping and clearance: the two weeks you can compress

Sea freight to Europe or the US runs 28 to 35 days door to door. Air freight is five to seven days and costs roughly $6.50 to $9.00 a kilogram, which on 42 kilograms is real money.

What you can compress is the paperwork. A commercial invoice with the wrong HS code, or a packing list that does not match the carton count, can hold a shipment at destination for a week. That is not the factory's delay, but it is your calendar.

How to pull four weeks out of a small run

Four moves, in order of effect. Approve the lab dip within 24 hours, and approve the round-one comments in one pass with numbers rather than adjectives.

Then book stock fabric instead of dyed-to-order, and give us a three-week delivery window instead of a fixed date. Those last two are the ones brands resist, and they are worth the most.

Do all four and a twelve-week run can land in nine. Do none of them and sixteen weeks is not bad luck, it is arithmetic.

In 2021 we quoted nine weeks on a 60-piece run and delivered ten weeks and two days. We had not built any float around the second mock-up round, and the client had a stockist appointment. The number we quote now includes the float, which is why it reads twelve rather than nine.

Our cut and sew page sets out the four stations and what each one needs from you. The FAQ page answers lead time questions, and the case studies page lists delivered runs with their actual dates. Our production process page shows where each of the four blocks starts and ends. The Fashion Pains Solver we are bringing online will map your target launch date back to a deposit date and flag the blocks that cannot absorb a slip.

A private label clothing manufacturer wins the order at the sample and loses it in bulk. We have run the same style from sample to shipping hundreds of times, and the failures sit almost always in the gap between an approved sample and a production line that has never seen it.

So what happens after you approve? You pay 30 percent, we buy the fabric and cut, and 70 percent is due against the shipping documents.

Between those two payments sit three inspections, a bulk fabric rest, and one dress-form fitting per size. The sample is a promise. Bulk is where it gets kept or broken.

What happens between sample approval and the cutting table

The approved sample becomes the sealed sample, and it stays in the sample room for the whole run. Every operator works against it, not against a photo.

The sealed sample is signed by both sides before cutting starts. If there is a dispute at the end, the conversation is then about one garment rather than about memory.

Two documents travel with it. The spec sheet, with a tolerance on every point of measure, and the trim card, with the actual labels, zips and thread from the sample. If a point of measure has no tolerance, nobody can fail it, and nobody can fix it either.

The fabric is then rested before cutting. Twenty-four hours minimum for a woven, longer for anything with stretch.

What the 30/70 payment actually buys

The 30 percent covers the fabric and the trims, which have to be paid for before they leave the mill. That is the whole reason the deposit exists, and it is not a trust exercise.

The 70 percent is due against the shipping documents: the commercial invoice, the packing list and the inspection report. Not against a promise, and not before the cartons are counted.

We will not ask for more than 50 percent up front, and we will not take payment into a personal account. Both are worth walking away from.

What we need from you during the four weeks

Three things, and all of them are yours to supply rather than ours to chase.

Final label and care artwork as vector files, not a screenshot. The care text in every language of the market you are selling into, because a French label on a garment sold in Germany is a compliance problem rather than a design choice.

And any change request in writing before cutting starts. After the lay is cut, a change is a re-cut, and a re-cut on 30 pieces is a real cost rather than a small one.

Three inspections, and what each one catches

Inline, during sewing. This is the cheapest place to catch a problem, and it catches tension, seam slippage and the first sign that a size is drifting.

End of line, on the finished garment. Measurements against the spec sheet, symmetry, and the fit on a form.

Pre-shipment, on packed cartons. Piece count, colour consistency across the run, and the label and care text one last time.

We use AQL 2.5 for major defects on bulk, which allows a defined number of minor issues rather than pretending zero is achievable.

The report comes to you with the defect count, not just a pass or fail. If the number is inside the limit we ship and tell you what it was. If it is outside, we fix the run and the schedule moves.

Dress form fitting: the check a tape measure misses

Every size is fitted on a form before packing, with the approved sample fitted beside it. A garment can measure perfectly and still hang wrong.

That is not theory. In 2020 we shipped 40 dresses that all measured inside tolerance and all hung short at the hem, because the sample fabric had been rested and the bulk had not. The measurements passed and the garment failed. Bulk cloth now rests a full day before cutting, and it is written into our process.

Where a bulk run goes wrong after a perfect sample

Three places. A substitute trim that looks identical in a photo. A second dye lot that reads one shade warmer under daylight. And a correction made to the sample after approval that never reached the spec sheet.

All three are avoidable with one habit: put the change in writing on the spec sheet, not in a chat message. We have been on both sides of that mistake, which is why the spec sheet is now the only document that counts.

There is a fourth, less dramatic one: a run that is rushed to catch a date. Compression comes out of the rework window rather than out of the sewing, and a run with no rework window ships whatever it has.

What you get at shipping

Cartons packed by size and colour, a packing list that matches the carton count, the inspection report, and the trim card with supplier reference numbers.

The trim card matters more than it looks. It is the only thing standing between a matching label on your reorder and a 500-piece minimum from a trim supplier who has never heard of you.

The balance falls due when those documents are in your hands, not when the ship leaves. That gives you a clear look at the run before the last payment moves.

Our private label page sets out what we produce under your brand from 30 pieces. The FAQ page answers deposit and inspection questions, and the case studies page lists delivered runs with their inspection outcomes. Our production process page shows where the 30/70 payments and the three inspections sit. The Fashion Pains Solver we are bringing online will compare the approved sample's spec sheet against the bulk inspection log and flag drift before the cartons close.

A private label clothing manufacturer is paid through a bank, not a chat app. Before the first transfer leaves your account, four things decide whether the money arrives, arrives whole, and lands in the right company's name.

So how do you actually pay a factory in Guangzhou? A bank transfer against a proforma invoice, split 30 percent at approval and 70 percent against the shipping documents. That is the standard, and almost everything else you have read about is either a fee trap or an unnecessary layer.

We have been paid every way a small brand can think of, and the failures are always the same three. Here is what we ask clients to do, and what we refuse.

What payment methods a factory really accepts

Telegraphic transfer is the norm, and it is what our bank account is built for. We invoice in US dollars or renminbi, whichever suits your books, and the beneficiary name is the company on your contract.

Letters of credit exist, and they are the wrong tool below about ten thousand dollars. The bank charges run one hundred and fifty to four hundred dollars, and the document checking adds a week to your schedule. For a thirty-piece first order that is a large tax on a small invoice.

PayPal and card payments get offered by some agents. They work for sample fees, and they fall apart on bulk, because the fees and the dispute window are built for consumers rather than factories. Wise is cheap until a receiving bank decides it does not like the reference, which happens more often than the marketing suggests.

WeChat Pay and Alipay are fine for a sample fee of a few hundred dollars. We will not take a bulk order that way, and you should not offer.

The proforma invoice is the document that matters

A proforma invoice is not a receipt and it is not a contract. It is the payment instruction, and it carries five things you must check line by line.

• The full legal company name, spelled exactly as it appears on the bank account.
• The bank name, branch, account number and SWIFT code.
• The amount, the currency, and the incoterm.
• The style and quantity it is paying for.
• The date, so you know which version is current.

If the account name and the company name do not match, stop. A mismatch is the single most common sign of an agent passing your order to a factory you have never spoken to.

Why we date every invoice

In 2020 a client paid an invoice we had replaced three weeks earlier, because the old bank details were still in their inbox. The transfer bounced, sat in an intermediary bank for eleven days, and came back lighter by sixty-eight dollars in fees. The proforma now carries a date, and the date is the version that counts.

The 30/70 split, and where each payment lands

Thirty percent at sample approval covers the fabric and the trims, which have to be paid for before they leave the mill. That is the whole reason the deposit exists, and it is not a test of trust.

Seventy percent falls due against the shipping documents: the commercial invoice, the packing list and the inspection report. Not against a photo of cartons, and not before the count matches.

Two rules we hold to. We never ask for more than fifty percent up front, and we never take payment into a personal account. If a supplier asks for either, the risk is not worth the discount you think you are getting.

What happens if you cancel

Before fabric is cut, the deposit is refundable minus what has already been spent, and we say which lines those are in writing. After the lay is cut, the fabric is yours and the deposit is spent, and no factory can give it back without taking the loss themselves.

Fees and currency: the money that disappears

A wire is not free. Sending a bank charge of twenty-five to forty-five dollars is normal, and an intermediary bank can quietly deduct another fifteen to fifty dollars on the way through.

Ask your bank for the split. Sending the fee with the payment, rather than letting the receiving bank take it, means the amount we receive matches the invoice. When it does not, we have to raise a short-payment note, and the order waits a day.

Quoting in dollars or in renminbi is a real choice. If your bank gives you a good rate on renminbi, take it, because it removes one conversion from the chain. If it does not, pay in dollars and let us carry the conversion on our side.

Three payment mistakes we have watched

• Paying a personal account to save a bank fee, then having no invoice, no receipt and no recourse when the order slips.
• Transferring before the proforma invoice exists, so the reference number does not match anything and the payment is held as unidentified.
• Paying in full up front to "secure a production slot". A slot that only exists if you pay everything is not a slot worth holding.

Our private label page sets out what we produce under your brand from 30 pieces. The FAQ page answers deposit and invoice questions, and the case studies page lists delivered runs with their payment terms. Our production process page shows where the two payments sit against sampling and bulk.

A small batch clothing manufacturer ships thirty pieces in two or three cartons, not a container. That single fact changes every freight number you were quoted, and it is why a quote built for five hundred pieces reads like nonsense at your volume.

So which way should a first order go? Express for anything under about twenty-five kilograms, air freight once you pass forty-five kilograms, and sea only when the cartons fill more than a cubic metre or the calendar gives you five spare weeks.

Here is the arithmetic we run with clients, using real cartons rather than rounded ones.

How many cartons thirty pieces actually fill

A folded dress takes about 0.06 to 0.09 of a cubic metre per carton, and a carton holds fifteen to twenty pieces. Thirty dresses is two cartons and roughly eighteen kilograms.

A lined blazer is bulkier. Thirty of them run to three or four cartons at 0.12 of a cubic metre each, and the weight crosses twenty kilograms on its own.

Those numbers matter because freight is priced on the greater of actual weight and volumetric weight, and volumetric weight is calculated by dividing the carton's cubic centimetres by five thousand. A light, bulky carton can bill as heavier than it is.

Express, air and sea, side by side

Express, meaning DHL or FedEx, runs three to five days door to door. On an eighteen-kilogram shipment to the United States or Europe, expect two hundred and twenty to three hundred and twenty dollars, all in.

Air freight takes five to seven days and prices at six dollars fifty to nine dollars per kilogram, plus eighty to one hundred and fifty dollars in handling and clearance. Most forwarders set a minimum around forty-five kilograms, so a small shipment pays for weight it does not have.

Sea freight, less than container load, takes twenty-eight to thirty-five days. The ocean leg looks cheap at one hundred and twenty to two hundred dollars for half a cubic metre, but destination charges of one hundred and fifty to three hundred dollars land on top, and they do not care how small your shipment is.

What the three options cost on one real order

Take thirty viscose dresses, two cartons, eighteen kilograms, 0.14 of a cubic metre. Express comes to about two hundred and sixty dollars. Air freight hits its minimum and quotes around three hundred and forty. Sea lands near three hundred once the destination fees are counted, and it arrives a month later.

Express wins on that order, and it is not close. The twenty dollars you think you save by sea buys you four weeks of waiting.

When sea becomes the right answer

Past about one cubic metre, or when the order is a reorder with no launch date attached, the ocean rate starts to win. Fifty blazers is the point where we usually suggest pricing all three and letting the calendar decide.

Port to port, or door to door

FOB means the freight becomes your problem at the port, and you book your own forwarder. DDP means it stays ours until the cartons reach your address, duty paid.

For a first order, door to door removes a whole set of decisions you are not yet equipped to make. For a reorder that ships twice a year, booking your own freight is worth the effort and usually saves a few percent.

The paperwork that travels with the cartons

Every shipment needs a commercial invoice, a packing list, and either an air waybill or a bill of lading. A certificate of origin helps at customs where a trade agreement applies.

The packing list is the one to read carefully. It should break the cartons down by size and colour, because that is the document you count against, and the place a shortage shows up first.

We send the pre-shipment inspection report with the same email. It is the record that the garments inside those cartons were measured before the lid went on.

How we pack a small order

Every garment goes into a polybag with the size on a sticker, and the cartons carry the style, size range and carton number. That marking is what makes the count on the packing list verifiable at the other end.

Cartons are sized to the shipment rather than to a pallet, because a small order is handled by people and not by a forklift. Thirty dresses usually fit in two cartons we can carry by hand.

If your garments need to arrive shelf-ready, say so at the quote. Adding a hangtag, a folded neck label and a branded bag is a packing line task, and it adds two to three days to the schedule and a small per-piece cost.

Insurance, and what happens in transit

Express and air shipments are insured by the carrier to a declared value, usually a low default per kilogram. If the order is worth more than that, declare the real value or buy a top-up, because the default will not cover a thirty-piece run at retail.

Sea freight is insured separately through the forwarder. Ask for the premium before you choose the mode, since it is small next to the cost of losing a launch.

If a carton arrives damaged, photograph it before opening, keep the packaging, and file with the carrier inside seven days. The factory cannot claim on your behalf for a box we no longer control.

Two shipping mistakes we have seen

A client booked their own courier for a sixty-kilogram air shipment and paid nearly double our rate, because they had no account discount. Ask your factory for the freight quote even if you plan to arrange it yourself.

In 2021 we shipped forty dresses by sea to save a client one hundred and eighty dollars. They arrived thirty-three days later, four days after a pop-up that had been booked for months. The saving cost the launch, and we now ask for the shelf date before we recommend a mode.

Our small batch page covers what we run from 30 pieces, and our cut and sew service covers the orders where you supply the cloth. The FAQ page answers freight and delivery questions, and the case studies page lists delivered runs with their shipping mode.

A high end clothing manufacturer is judged after the invoice, not before it. What happens when cartons arrive with a fault decides whether you order a second time, and it is the part most quotes never mention.

So what do you do when bulk turns up wrong? Open every carton within seventy-two hours, photograph the fault beside a tape measure, and send one table rather than forty messages. A claim made that way is settled in days. A claim made in a chat thread takes weeks.

We have paid for our own mistakes more than once, and this is the process we now use for every client, on every run.

The 72-hour window, and why it exists

The claim window starts when the cartons are delivered, not when you open them. Inspect within three days, because after that neither side can tell whether a fault left the workshop or happened in a warehouse.

The baseline is the pre-shipment inspection report that travelled with the shipment. It records the defect count and the measurements before the lid went on, which means a claim becomes a comparison rather than an argument.

Why three days and not thirty

A month is enough time for a garment to be worn, washed, altered or sold, and after that the evidence is gone. Three days is long enough to open every carton on a small order, and short enough to still be talking about the same run.

What counts as a defect, and what is a tolerance

A defect is something outside the spec you approved. A measurement outside its tolerance, a seam that slips under tension, a shade band across a panel, the wrong trim, a care label in the wrong language.

Not a defect: a measurement inside the tolerance band, a colour inside the agreed grey scale, or a change you approved after the sample.

That distinction only works if the spec has numbers on it. In 2019 a client claimed a short hem on thirty pieces, and the spec sheet said nothing about hem length at all. We had no tolerance to measure against, so we settled it as goodwill and changed the sheet. Every point of measure now carries a tolerance, and the conversation is about a number rather than an opinion.

The three outcomes we offer

Repair, at our cost, when the fault is local and a tailor can fix it. A dropped hem or a loose button is a twenty-minute repair, and we would rather pay for that than remake a garment.

Remake, when a batch is out of spec in a way a repair cannot hide. On a thirty-piece run, a remake of four pieces takes five to seven days, and we hold the fabric to do it.

Credit, when the fault is cosmetic and you would rather sell the pieces with a note. A credit note against your next order is often the fastest answer, and it keeps the garments out of a landfill.

Who pays the freight

If the fault is ours, we cover the return freight and the replacement freight. If a repair can be done at your end by a local tailor, we pay the tailoring bill, which is almost always cheaper than shipping a garment eight thousand kilometres twice.

What we will not claim on

Fabric you supplied that shrinks or shades differently from its own test report. A deviation you approved in writing. Wear that happened after delivery. And anything reported after the seventy-two-hour window with no inspection record behind it.

How a claim is priced

A claim is priced on units, not on the order. If four of thirty pieces are wrong, the claim is four pieces, and the number we settle on is the unit cost rather than the retail price.

Where a garment is sellable after a small fix, we usually offer the tailoring cost plus a credit rather than a remake. It is faster for you and cheaper for us, and the garment still reaches a customer.

We put the outcome in writing on the same day we agree it, with the quantity, the remedy and the date it will be done. A verbal agreement on a call is not a claim record.

What a claim teaches the next run

Every claim we settle changes something upstream. A tolerance that was missing gets added to the spec sheet. A trim that shaded differently gets its supplier number written onto the card. A seam that slipped gets a stitch density on the construction sheet.

That is the difference between a factory that fixes a fault and one that stops repeating it. Ask any supplier what their last claim changed, and listen to whether the answer is a document or an apology.

How to make a claim that gets answered

Send one email with one table. Style, size, quantity affected, a photograph of the fault beside a tape measure, and the measurement against the spec.

Do not open a chat thread with forty photographs and no summary. The factory reads the table, and the photographs confirm it. A thread is read by whoever is on shift.

Ask for a decision within five working days, and say what outcome you want. Repair, remake or credit. A clear ask gets a clear answer.

Why the inspection report protects both sides

It stops a factory from denying a fault that is real, and it stops a brand from claiming on a garment that measured correctly when it shipped. A high quality factory treats that report as the start of the conversation, not the end of it.

Our high end production page sets out the standards we hold and the inspections behind them. The FAQ page answers quality and claim questions, and the case studies page lists delivered runs with their inspection outcomes. Our production process page shows where the three inspections sit.

A small batch clothing manufacturer quotes fabric by the metre, but the number that decides how a garment hangs on the body is grams per square metre. Pick the wrong weight and thirty perfectly sewn pieces still feel wrong on the rail.

So how do you choose a weight before you order? Work backwards from the silhouette.

A fluid slip dress wants 90 to 130 GSM viscose. A structured midi wants 180 to 260 GSM. Suiting sits at 240 to 280 GSM. Put the number on the tech pack, because "medium weight" means nothing to a mill.

Weight is the cheapest thing to fix before cutting and the most expensive thing to fix after. This is how we think about it on every quote.

What GSM measures, and what it does not

GSM is grams per square metre: the mass of one square metre of fabric. It tells you how dense and heavy a cloth is, which is a fair proxy for how it will drape and how opaque it will be.

It does not tell you everything. Two fabrics at the same GSM can behave nothing alike. A loose 140 GSM linen and a tight 140 GSM TR suiting weigh the same and fall completely differently, because weave and yarn count matter as much as weight.

Where weight beats every other spec

On a first order, weight is the fastest way to predict the two failures clients hate most: a garment that is see-through, and a garment that will not hold a line.

Fabric weight by garment type: our working ranges

These are the ranges we quote most often for the categories we actually make.

• Slip dresses and bias-cut skirts in viscose or satin: 90 to 130 GSM.
• Shirts and blouses in cotton or cotton-linen: 110 to 160 GSM.
• Structured dresses and pencil skirts in TR suiting: 180 to 260 GSM.
• Trousers and suiting in TR or poly-viscose: 240 to 280 GSM.
• Summer linens: 160 to 220 GSM, and they soften with every wash.

If you are between two ranges, ask for both swatches in the same colourway. The difference is obvious in the hand and invisible on a spec sheet.

Why weight and drape are not the same thing

A heavier fabric is not automatically a better one. On a bias-cut dress, extra weight can pull the hem out of shape and make a size 14 hang longer than a size 8. On a pleated skirt, too little weight means the pleats open as the wearer walks.

Drape also depends on the finish. A 130 GSM viscose with a soft finish and a 130 GSM viscose with a stiff finish are two different garments, and only one of them is the one you sketched.

The see-through test you can do yourself

Hold a swatch against a window in daylight and look for your hand. If you can read its outline through a white dress fabric, it is too light for that style, or it needs a lining.

What changing weight does to your cost per piece

Weight moves price more than most people expect, because mills sell by weight as often as by metre. Moving a dress from 110 to 150 GSM can add 15 to 25 per cent to the fabric line before you change anything else.

On a thirty-piece run that is real money. A heavier cloth also uses slightly more metres, because it is thicker and the marker lays a touch deeper.

Where it is worth paying

Pay for weight where the client will feel it: a trouser that holds a crease, a dress that skips the lining. Save on weight where nobody touches it, like the back panel of a top worn under something else.

Three weight mistakes we have made

In 2019 we quoted a 110 GSM viscose for a white slip dress because the client loved the colour on the swatch. The first mock-up was see-through under studio light. We re-sourced to 135 GSM, added a cupro lining, and the dress cost nine dollars more per piece than the original quote.

The lesson was not "buy heavier". It was "test the fabric for the garment, not on the table". We now run a light test on every white and pale fabric before it enters a quote.

A second one: we once approved a linen trouser at 150 GSM that creased into a mess after one wash, because the finish was not pre-shrunk. The weight was fine. The finish was wrong.

How to specify weight so a mill cannot guess

Write three things on the tech pack: the GSM, the tolerance, and the composition. A mill can hit 200 GSM in a dozen different blends, and each one drapes differently.

Ask for a shrinkage test before the bulk fabric is cut, and a lab dip for the colour. Weight and colour are separate problems, and solving one does not solve the other.

What "GSM plus or minus" should say

We quote a tolerance of plus or minus five per cent on weight. Anything wider and the garment you approve is not the garment you receive.

What we put in writing

Every quote from us lists the fabric with its GSM, composition and tolerance as its own line, so you can see what the weight is costing you.

Our small batch program covers runs from thirty pieces, and our FAQ page answers fabric and minimum questions. The case studies page lists delivered runs by fabric, and the production process page shows where fabric testing sits before cutting.

A high end clothing manufacturer is not a price on a page. It is a person who answers at 7am their time because that is your evening, and who tells you what they cannot do before you pay a deposit.

So what should you expect when you work across an ocean? Guangzhou runs on GMT+8.

From New York the usable window is your evening and their morning. From London it is your afternoon and their evening. Batch your questions for that window and a factory ten thousand kilometres away feels local.

We have run orders for clients in six time zones, and this is what we tell every new brand in week one.

The overlap window, by city

Here is the honest overlap, using a normal Guangzhou day of 9am to 6pm.

• New York, GMT-5: your 8pm to 11pm is their 9am to noon.
• Los Angeles, GMT-8: your 5pm to 8pm is their 9am to noon.
• London, GMT+1: your 8am to 11am is their 3pm to 6pm.
• Milan and Paris, GMT+2: your 9am to noon is their 3pm to 6pm.
• Dubai, GMT+4: your 8am to 1pm is their noon to 5pm.

If you write outside the window, do not expect an answer until it opens. That is not a snub. It is a person asleep.

Who actually replies, and when

On a real order you will talk to two or three people: a sales contact, a merchandiser who runs your order, and sometimes a technician who answers fit questions.

Ask who your day-to-day contact is and who covers when they are away. A factory with one overloaded contact goes quiet the week that person takes leave, and you find out mid-order.

What a working day looks like on their side

Your contact is often on the factory floor, not at a desk. Between 9am and 11am they are checking a lay or a fitting, which is exactly when your email lands. Expect a reply later the same day, not in five minutes.

What to send in your first email

Your first email sets the tone for the whole order. Send the style, the target quantity, the fabric if you have chosen one, and the date you need the goods in hand.

A factory can answer a specific brief in a day. A vague one takes a week of guessing, and the guess usually comes back wrong.

If you do not have a tech pack, say so and send reference photos with your measurements. We will build the pack, and the quote comes faster for it.

Email, WeChat or WhatsApp: what to use for what

Keep the record on email. Specs, revisions, prices and approvals belong in a thread you can search in eight weeks.

Use WeChat or WhatsApp for what they are good at: a quick photo of a mock-up, a yes or no on a colour, a two-line update. If a decision is made in chat, paste it back into the email so it is on the record.

Why we send one digest a day

In 2021 we lost five days on a Los Angeles order because the client asked fit questions at their 3pm, which is 6am in Guangzhou. By the time we opened the thread it was their night, and the answer waited another day. We now send a fixed daily digest at 6pm our time so nothing sits over a weekend.

Why the factory goes quiet at Chinese New Year

Chinese New Year stops the workshop for two to three weeks, and the sample room slows about two weeks before that. A factory that tells you this in October is being straight with you. One that promises a February sample run and then goes silent is not.

Plan any spring order around the holiday, and confirm the exact dates each year, because the holiday follows the lunar calendar and moves.

A reply time you can hold a factory to

Ask for two numbers before you pay a deposit: how fast a quote comes back, and how fast a question is answered during production.

We quote within 24 hours of a tech pack, and we answer production questions the same working day. If a question needs a sample room to check something, we say so and give a date.

The three-day rule for silence

If you hear nothing for three working days on an open question, chase it once in writing. If that gets no reply either, that is information about the factory, not about your email.

What slow replies usually mean

A slow reply does not always mean disinterest. It often means the person you wrote to is on the floor and your message is at the bottom of a long list.

The opposite is worth watching too. A factory that answers every message in two minutes, at any hour, may be a trading company passing your order to someone else. A workshop answers between other jobs, because it is actually making things.

Judge a factory by the quality of a slow reply, not the speed. A useful reply answers the question, flags a risk you did not ask about, and gives a date. A fast reply that dodges all three is worth less than a slow one that does all three.

What we put in writing

Every quote from us names your contact, their working hours, and the reply time we hold ourselves to. Our high end production page sets out the standards behind those promises.

The FAQ page answers lead-time and communication questions, the case studies page lists runs by city, and the production process page shows where each approval sits.

A private label clothing manufacturer will tell you the fabric minimum before you order. Almost none will tell you what happens to the fabric you pay for and never cut.

So where does the leftover go? If we bought it, it stays with us and we can hold it for your reorder or send it to you at your cost. If you bought it, it is yours the moment it lands, and you decide. The number that matters is how much you will actually have left.

On a thirty-piece dress run you buy about 100 to 120 metres and cut 85 to 95 of them. The rest is marker waste, end-of-roll and spares. Here is how we plan for it.

How much fabric is actually left over

On a small run, marker waste runs 8 to 12 per cent. At high volume it drops to 4 to 6 per cent, because a marker can nest more sizes across a wider lay.

On 120 metres, that waste is 10 to 14 metres. Add the end of the roll, which mills often ship three to five per cent over, and a small order can leave 15 metres on the shelf.

Why small runs waste more

A thirty-piece order has fewer sizes to nest, so the marker cannot share fabric between a size 8 and a size 16. The gaps between pieces are wider, and wider gaps are wasted cloth.

Ask the factory for the expected leftover in metres, not as a percentage. A number in metres is something you can plan a second style around; a percentage is a number you cannot act on.

Marker waste: the 8 to 12 per cent nobody quotes

Most quotes show a fabric cost and stop. The marker percentage is where a small run quietly costs more than a large one, and it is worth asking for the number.

If a factory tells you the marker efficiency, you can sanity-check the fabric line yourself. If it cannot, it may not be measuring it.

Who owns the leftover, and who stores it

Whoever paid for the fabric owns it. If you supplied the cloth, the leftover is yours and we will pack it back with the shipment if you ask, at your freight cost.

If we sourced it, we hold it. For orders with a likely reorder we store the leftover free for twelve months, tagged with the style, colour and dye lot. After that we either ship it or, with your written go-ahead, recycle it.

Ask this before the deposit

Ask where your leftover will be stored, for how long, and at what cost. A factory that will not answer that is a factory that will not remember your dye lot next season.

Three ways to use leftover instead of shelving it

The cheapest fabric you will ever buy is the fabric already in the building. Here is how we help clients use it.

• Cut a second style from the same roll in the same colour, so two looks share one fabric line.
• Run a small add-on style, like a simple skirt or top, to round out a capsule.
• Save it for next season's sample rounds, so mock-ups do not eat new cloth.

All three only work if the leftover is real, matching, single dye lot fabric. Mixed offcuts are not useful for production, only for testing.

How to plan an order so you waste less

Waste is not fixed. Three habits cut the leftover on a small run before a single metre is bought.

• Nest five sizes in one marker instead of splitting by colour, so the gaps close.
• Buy one colour across two styles, so the roll is used down to the last usable length.
• Add a simple second style that can absorb the offcuts of the first.

None of these need a bigger order. They need the plan set before the fabric is cut, because once the lay is on the table the leftover is already decided.

What dead stock really costs

Dead stock is not free just because it is already paid for. It is capital sitting on a shelf, and it is a risk that the next dye lot will not match it.

In 2022 a client bought 300 metres of dyed-to-order fabric for a 40-piece run. About 240 metres stayed on our rack for fourteen months. When they reordered, the new dye lot came in 1.5 Delta E away from the old one, and the two runs did not match side by side on the rail. They sold the old pieces as a separate drop.

The two costs people forget

Storage and dye-lot drift. Neither shows on the invoice, and both show up eighteen months later when you try to reorder.

When to donate, and when to keep

Keep leftover that matches a style you may reorder. Donate or recycle the rest, because a shelf of unmatched offcuts is a cost with no upside.

Guangzhou has textile recyclers who respin clean cotton and viscose. We will send dead stock there with your written go-ahead, and we do not sell your fabric to anyone.

What we put in writing

Every quote from us states the estimated leftover, who owns it, and the storage terms, so the question never comes up at shipment. Our private label program covers fabric sourcing and storage as separate lines.

The FAQ page answers fabric and minimum questions, the case studies page lists runs by fabric, and the production process page shows where fabric is received and stored.

A high end clothing manufacturer should be able to name the grade of the goods it ships. Ask most factories and you get a shrug, because grading is a habit few shops ever build. We grade every run A, B or C, and we tell you which one you are buying before the deposit clears.

What separates the grades? Not the fabric and not the price. It is the number of defects we allow per hundred pieces, and where those defects are allowed to sit.

An A run tolerates almost nothing visible. A B run tolerates small things a shopper will never notice on a hanger. That is the whole idea, and it is worth settling before you order, not after the cartons land.

Here is how we draw the line, and what each grade means in practice.

What the three grades actually measure

A grade is not a feeling. We count two things: defects per hundred pieces, and how serious each one is. A skipped stitch inside a facing is not the same as a broken stitch across a front seam, even though both count as one defect on paper.

So we split defects into three levels. Critical means the garment cannot be sold or worn. Major means it is visible and a shopper would notice. Minor means it shows only under a light, to someone inspecting it on purpose.

Every grade starts with zero critical. That part is not negotiable at any price.

A grade

Zero critical, under two major per hundred pieces, and under three minor per hundred. On a thirty-piece run that usually means nobody finds anything on the rail. This is the grade we ship when a brand tells us their buyer inspects before placing the next order.

A grade costs more, because it wastes more. If we cut thirty and five do not meet the line, we remake five. On a small run that is a real number, and it shows up in the per-piece price.

B grade

Zero critical, under four major per hundred, minor defects up to eight per hundred. A B run is a solid commercial garment. What makes it B are the things you find only when you look: a slightly uneven topstitch, a small shade difference between two panels, a button a millimetre off centre.

C grade

Zero critical, major defects up to eight per hundred. C is the grade for test markets, sample sales and stock that sells on price. We will make it, but we will not call it A, and we will not let a client discover the difference on a customer's doorstep.

Why most factories never publish a grade

Because a grade is a promise, and a promise can be checked. If a factory writes "A grade" on a quote, you can hold it to a defect count. Most would rather leave the words out and settle disputes one at a time.

We put the grade on the quote for the same reason we write the deposit split. It removes an argument later, when the goods are already on a boat.

What to ask for in writing

Ask three things. The grade, the defect count per hundred that defines it, and what happens when a piece fails. A factory that can answer all three is measuring its own output. One that cannot is guessing.

Ask who pays for a remake, too. Our answer is simple: if the defect is ours, we absorb it. We inspect three times before shipping, so a defect that reaches you is a defect we should have caught.

How the grade changes your price

Grade is a lever, not a fixed cost. Moving a simple dress from A to B might save eight to twelve per cent, because we stop remaking pieces that only a light would expose.

On a structured blazer the gap is wider, because rework is slower. We have quoted the same style at $92 in A and $79 in B, and the client took A because their stockist checks every piece.

If you are testing a market, B is a fair choice. If you are pitching a stockist, do not gamble on B. One bad piece in a showroom costs more than the difference.

A grade mistake we made

In 2021 we shipped a forty-piece run we called A without ever writing down what A meant. The client counted four minor defects and asked for a discount. They were right, because we had never defined the word.

We now write the count on every quote. It cost us a hard week, and it taught us that a grade without a number is just a word.

Where grading sits in your order

Grading is not a separate step. It happens during the bulk run, where we inspect inline, at end of line, and again before shipment. The production process page shows all three checkpoints.

If you want to see grades on real orders, the case studies page lists runs by size and fabric. The FAQ page answers the questions clients ask most.

Our high end manufacturing program is where the A grade lives, and it is where most first orders start. We are also building a Fashion Pains Solver so brands can log a defect and see how we resolved it; it is still in training, so for now that log runs through your account manager.

A small batch clothing manufacturer does not buy fabric whenever you call. We buy inside a window, and the window changes what you pay and what you can get. Plan a spring dress in October and you are early; plan it in March and you are competing with every other brand for the same roll.

So what actually changes between spring and fall? Three things: the mill calendar, the fabric weight the market wants, and how long a dye lot stays available. None of them are about your order size, and all of them can move your price by ten to twenty per cent.

Here is how we plan a small run around a season, and where clients lose money by ignoring it.

Why the same fabric costs more in one season

Mills run on their own rhythm. A mill that weaves linen for spring has it ready in winter and sells out by early spring. Ask for the same linen in June and you are buying from what is left, or paying for a fresh dye run at a minimum you cannot reach on thirty pieces.

Lightweight viscose and cotton blends follow the same curve. The cloth is cheapest when it is in season to weave, not when it is in season to sell.

The mill calendar nobody shares

Chinese mills close for Spring Festival, and many shut a week around National Day in early October. A dye house that normally takes seven days can take three weeks across the holiday.

If you need spring fabric, we book it in November or December. If you need fall fabric, we book it in May or June. Anything later and you are asking a mill to squeeze you into a queue that is already full.

Ask your factory when they place their fabric orders, not just when they need the cloth. A factory that buys late is a factory that will be late.

Spring versus fall: what changes in the cloth

Spring and summer styles lean light: 90 to 130 grams per square metre for a slip dress, cotton poplin around 110, viscose challis near 120. Fall styles lean heavier: crepe and suiting at 180 to 260, and structured skirts that need body.

Heavier cloth behaves differently in the marker and in the making. It shrinks less in the wash but presses slower, so the finishing room takes longer. That is a lead-time fact, not just a price one.

We quote the weight and the tolerance together, because a "light viscose" means nothing without a number and a plus or minus five per cent.

How we plan a 30-piece order around a season

Thirty pieces needs roughly 100 to 120 metres, depending on the style. That is below most dyed-to-order minimums, which sit around 300 metres, so on a first run we usually buy stock fabric rather than dye our own.

Stock fabric is cheaper and faster, but it comes in the colours and weights the mill already runs. If your brand colour is unusual, you either accept the closest stock shade or you pay for a dye lot and shelve the extra metres.

Either way, we confirm the fabric before we confirm the delivery date. A date set before the cloth is locked is a guess.

Dye lots and the cost of waiting

Every dye lot is slightly different. If you buy thirty metres now and thirty next season, the two may not match side by side. Delta E is the number that measures it, and a gap of 1.5 is often visible on a rail.

So we plan a season's styles on one lot where we can. Two styles, one fabric, one lot, and the leftover from the first feeds the second.

A season we got wrong

In 2022 a client asked for a summer capsule in April. We said yes and bought stock viscose left over from the spring weave. It arrived two shades deeper than the swatch, because the swatch had come from an earlier lot.

We remade the sample twice and lost nine days. Now we photograph the actual roll we will cut, and we send the lot number before we lay a single piece.

What to lock before you order

Lock four things: the fabric and its weight, the lot, the price per metre, and the season you intend to sell in. Everything else can move.

If you want to see how we handle fabric minimums on small runs, the small batch program explains the numbers. The case studies page lists runs by season and fabric, and the FAQ page covers the sourcing questions we hear most.

The production process page shows where fabric is received, tested and stored.

A cut and sew manufacturer will usually tell you the minimum, then go quiet on how many pieces it can make in a day. That second number matters more, because it decides your delivery date. A factory that says yes to fifty pieces and runs one slow line will still be late.

So how many pieces does a line sew in a day? For a simple dress with six to eight operations, a six-person line makes 40 to 70 pieces a day. A lined blazer with thirty parts on the same line makes 12 to 20. The order size rarely changes those numbers; the style does.

Here is how we work out daily output, and how to read a factory's capacity claim without getting fooled.

What actually decides daily output

Three things set the pace: how many operations the garment needs, how many people work the line, and how long the slowest operation takes. A line moves as fast as its slowest station, not its fastest.

If one person sets sleeves and each set takes four minutes, that station makes fifteen pieces an hour at best. No amount of people downstream changes it.

So we plan around the bottleneck station, and we tell clients which one it is.

Real numbers by garment type

On a six-person line, a simple slip dress or top runs 40 to 70 pieces a day. A structured skirt with a zip and lining runs 30 to 45. A lined blazer runs 12 to 20, because the canvas, the collar and the sleeves are hand-guided.

Those are working numbers, not marketing ones. They assume the fabric is on site, the sample is approved, and nobody is waiting on a trim.

If a factory quotes you a blazer at 60 a day on one small line, ask how many people work it. The answer will explain the number.

Why order size is not the bottleneck

Clients often think a bigger order gets made faster. It does not, up to a point. Fifty pieces and two hundred pieces both run at the same daily rate; the bigger order simply takes more days.

What a bigger order can do is justify a second line or a longer shift. That is a scheduling choice, and it is worth asking about.

For a first run, the calendar matters more than the count. We slot small orders between larger ones, so a fifty-piece run can move quickly if it lands in an open window.

How we schedule a 50-piece run

Fifty pieces on a six-person line is roughly one to two days of sewing, depending on the style. But sewing is not the whole run. Cutting, bundling, pressing, inspecting and packing each take their own slice.

In practice, a fifty-piece order spends about four to five working days on the floor after the fabric is cut. The rest of the twelve-week timeline is quoting, fabric, samples and shipping.

We give clients the floor days separately, so they can see where the time really goes.

Reading a factory's capacity claim

A factory that claims ten thousand pieces a month is telling you its ceiling, not its normal speed. Ask instead how many lines it runs, how many people per line, and how many orders are open right now.

A shop with three lines and eight open orders may be slower for you than a smaller shop with one open order. Capacity is not a number; it is a number minus what is already booked.

Ask what they would refuse in order to take yours. A factory that never says no is a factory that will delay someone, and it might be you.

The capacity mistake we made

In 2020 we accepted two fifty-piece orders in the same week and put both on one line. The changeover between styles cost us most of a day, and the second order shipped four days late.

Now we hold one style per line and stagger small runs. It looks less efficient on paper and it ships on time, which is the only efficiency that matters.

What to ask before you book a slot

Ask four things: daily output for your style, the bottleneck operation, the floor days after cutting, and the window you are being slotted into. If the answers are vague, the delivery date will be too.

If you are deciding between models, the cut and sew program explains what we take on and what you supply. The case studies page shows real run sizes and dates, and the FAQ page covers scheduling questions.

The production process page shows where cutting, sewing and inspection sit in the timeline.

A small batch clothing manufacturer can cut thirty pieces and still lose the whole run to one roll of fabric that behaves differently. We test fabric before we cut for exactly that reason, and the tests cost very little next to remaking an order.

So what should you test before production? Three things: shrinkage, colourfastness, and pilling. Shrinkage decides final size, colourfastness decides whether the colour survives a wash and a rub, and pilling decides whether the surface still looks new after a week of wear.

If a supplier will not share those three results, you are buying a guess.

Here is what each test measures, the numbers we work to, and how to ask for the paperwork without sounding difficult.

Shrinkage and colourfastness: why three tests decide a run

Fabric moves. Viscose can shrink 4 to 6 per cent on the first wash, linen can lose 2 to 3, and a stable polyester blend may lose under 1.

None of that is a defect; it is a property you plan for.

If you cut to final size and the fabric then shrinks, the garment arrives short. On a thirty-piece run there is no spare to absorb it, so every piece is affected, not a few.

The fix is simple: wash a sample length first, measure it, and cut with the shrinkage built into the pattern.

Test one: shrinkage after a home wash

We cut a 50cm by 50cm square, wash it at 30 degrees, dry it flat, and measure again. The change tells us how much to add to the pattern before the marker is made.

A fabric that shrinks 5 per cent needs roughly 5 per cent added to length and width. Skip it and a 100cm dress quietly becomes 95cm, which is a whole size at the hem.

We ask for a shrinkage figure on the fabric spec, with a tolerance. If it is missing, we test it ourselves and tell you the number before we cut.

Test two: colourfastness to washing and rubbing

Colourfastness is graded 1 to 5, where 5 means no change. We want 4 or better for washing, and 4 for rubbing, both dry and wet.

A dark viscose that grades 3 will bleed onto a pale lining in the first wash. On a dress with a contrasting panel, that shows up immediately.

If your style mixes a deep colour with a light one, ask for the rubbing grade specifically. That is the one that ruins a two-tone garment, and it is often the one nobody checks.

Light fastness matters too if the garment will sit in a shop window. We ask for a grade 4 for light when a style is likely to be displayed for weeks.

Test three: pilling and surface wear

Pilling is the small balls of fibre that form where the fabric rubs. It is graded 1 to 5 as well, and we want 4 or better for anything worn against a bag strap, a seat or a coat.

Blends pill more than tight weaves. A soft viscose-cotton blend can pill at 3 while the same weight in a tighter weave holds at 4.

Pilling is not a make defect. It is a fabric choice, and it is far cheaper to change fabric than to argue about it after delivery.

The pass numbers we hold a fabric to

Our working minimums: shrinkage within 3 per cent for woven fabric, colourfastness 4 for wash and 4 for rub, pilling 4 for any outer surface. Anything below that we flag before the order is confirmed.

Those numbers are not universal. Swimwear, workwear and childrenswear carry stricter standards, and we say so when a brief falls into those groups rather than pretend one number fits everything.

If your fabric fails one test, we do not bury it. We send the number and the two options: change the fabric, or accept the result and adjust the pattern to compensate.

We keep the results with the order file. If a question comes up two years later, we can show what the fabric tested at when we bought it, not what we remember.

How to ask a factory for fabric test reports

Ask for three things: the shrinkage figure after one wash, the colourfastness grades for wash and rub, and the pilling grade. Ask for the mill's report number and date, so the result can be traced to a lot.

A report without a date is not useful. Fabric from a new dye lot gets tested again, because the same code from a different lot can behave differently even when the composition matches.

If the only report is more than a year old, ask whether the mill has changed its finishing line since. Finishing is what usually moves the shrinkage number.

If the supplier only has a generic sheet for the fabric family, treat it as a starting point, not a guarantee for your roll.

The test we once skipped

In 2019 we took a client's own white viscose and cut a thirty-piece run without a wash test. The fabric shrank 6 per cent and the dresses came out short at the hem.

We remade the order at our cost. Since then no fabric is cut before a wash square is measured, even when the client supplies it and even when they say it is pre-shrunk.

If you are weighing up fabric options, the small batch program shows how we buy and test for short runs. The case studies page lists real orders and dates, and the FAQ page answers the common fabric questions.

The production process page shows where fabric testing sits, before cutting begins.

A cut and sew manufacturer that cannot tell you which bundle a panel came from cannot fix a fault. On a thirty-piece run there is nowhere to hide a mistake, so we number every panel before it leaves the cutting table.

How does a factory keep track of thirty or fifty pieces through a sewing line? With a cut ticket and a number on every panel. The ticket records the style, the size, the fabric lot and the count, and the number ties each panel back to that ticket.

Lose the number and you lose the ability to trace.

Here is how the system works, why the numbers matter, and what to ask a factory that says it does not need them.

What a cut ticket actually records

A cut ticket is a single sheet that follows one lay. It lists the style, the size breakdown, the fabric lot, the ply count and the number of pieces cut.

Every bundle carries a copy. When a fault appears on one dress, we read the ticket and know which roll and which lay it came from.

Without the ticket, a fault is just one dress with a problem, and we cannot tell whether it is alone or the first of twelve.

A ticket also carries the date and the cutter's name. If a fault appears three weeks later, those two lines tell us who to ask and when the lay ran.

Why panels are numbered by shade band

Fabric from one dye lot is not always one colour. Mills band it, and we cut within a band so a garment does not mix two shades.

The number on each panel includes the band, so a sleeve and a bodice from the same dress are always from the same shade.

On a deep colour this is the difference between a clean dress and one that looks patchy under shop light.

How bundles move through the line

Panels are bundled by size and by operation, then tied with a ticket. A bundle of thirty fronts moves together, so the sewer never picks up a piece from another order.

At the end of each operation the count is checked against the ticket. If a bundle arrives with nine pieces instead of ten, we stop and find the tenth before it becomes a missing garment.

That check is boring, and it is the reason we rarely lose a piece.

How many bundles a 30-piece run makes

Thirty pieces across five sizes, cut as one lay, becomes roughly thirty to forty bundles once split by size and by operation. Each bundle is small, often five to ten panels.

That sounds like a lot of paper for a small order. It is still less work than hunting for a missing sleeve at packing, and it is why a short run can be traced as tightly as a large one.

On a fifty-piece run the bundle count grows, but the system is identical. Size does not change the method; it only changes how many tickets you print.

What happens when a panel is lost

A panel can be cut wrong, stained, or simply misplaced. Because the ticket knows how much fabric the lay used, we know whether there is enough left to re-cut.

For a thirty-piece run the spare is thin. A re-cut often means buying a metre or two, and if the lot has moved on, the colour may not match.

That is why we tell clients early when a re-cut needs fabric, rather than quietly shipping twenty-nine pieces.

The re-cut rule we hold to

We re-cut from the same roll wherever possible. If the roll is gone, we cut from the nearest band and mark the piece, so you know which garment carries it.

We never swap in a panel from a different lot and hope nobody notices. That is the kind of shortcut that shows up two washes later.

If a re-cut will change the delivery date, we say so the day we find the problem.

We also log every re-cut on the ticket. If a run needs more than two, we tell the client, because a pattern of re-cuts usually points at a fabric or marker problem rather than bad luck.

What to ask a cut and sew partner

Ask four things: does every bundle carry a cut ticket, is the fabric banded before cutting, is the count checked after each operation, and how do you handle a re-cut?

A factory that answers all four in one breath has a system. A factory that says it does not need numbers is telling you it has never had to trace a fault.

You do not need to audit the system yourself. Ask for one cut ticket from your own order and read it. It should name your style and your lot.

If a factory has never shown a client a cut ticket, ask for one now. It is a two-minute request and it tells you more about their floor than a factory tour.

The bundle mix-up we had

In 2018 two fifty-piece orders in similar navy fabric shared a cutting table on the same day. One bundle of fronts was tied into the wrong order, and it was caught only at packing.

We now clear the table between orders and never cut two similar lots back to back. It costs us half a day and it stopped a repeat.

If you are choosing between production models, the cut and sew program explains what we take on and what you supply. The case studies page shows real run sizes and dates, and the FAQ page covers tracking and inspection questions.

The production process page shows where cutting and bundling sit in the timeline.

A private label clothing manufacturer usually puts a validity window on a quote, and clients often ignore it until they are ready to order. By then the fabric has moved and the number is gone.

How long should a quote last? Most garment quotes hold for 15 to 30 days, and the fabric portion is the part most likely to change. If a quote has no date and no window, it is not really a quote; it is an estimate.

Here is why prices move, what stays fixed, and how to hold a price long enough to make a decision.

Why a quote has a shelf life

A quote is built from today's fabric price, today's labour rate and today's exchange rate. Two of those three move on their own.

Fabric is the largest line on most quotes, often 30 to 45 per cent. When the mill moves, the quote has to move with it.

Labour and trims are more stable over a month, so the fabric line is where the risk sits.

What is locked and what is not

What we lock for the window: the sewing cost, the trim cost, the sampling fee and the margin. What can move: the fabric price and, if you pay in another currency, the exchange rate.

If you want a fully fixed price, we can do it by buying the fabric early and holding it. That converts a moving line into a fixed one, and it usually costs a small holding fee.

Ask which lines are fixed and which are open. A good quote says so on the page.

One more line worth pinning down: the shipping term. A quote can be FOB or DDP, and the difference on a small order is real. If the term is not written on the quote, you do not yet have a price for the goods in your hands.

Fabric prices move in three ways

First, the mill raises its base price, often when raw material or energy costs rise. Second, your quantity changes, so the price per metre changes with it. Third, the dye lot moves, and a new lot can carry a new price.

On a thirty-piece run you are usually buying from stock, so the base price is the one that moves.

None of these are hidden fees. They are the mill's real numbers, and they change for everyone.

There is a fourth, quieter move: the mill changes the minimum for that fabric. A cloth you could buy at 100 metres may need 300 next month, which changes the price per metre even when the base rate has not moved.

Currency and the exchange rate line

If you are quoted in USD and pay in EUR or GBP, the exchange rate sits inside the price. A two per cent move on a 3,000 dollar order is 60 dollars.

We quote in the currency you ask for, and we show the rate we used and its date. If the rate moves before you pay, we re-quote that line, not the whole order.

Clients who pay in the quote currency avoid this line entirely.

Does a sample fee expire too?

Sampling is quoted separately, and it usually holds longer than the bulk price because it does not depend on a big fabric buy. A sample fee can sit for two or three months.

What does expire is the credit. Many factories, including us, refund the sample fee against the bulk order if it is placed within 60 days. After that the fee stays a cost.

So if you are still deciding, ask how long the sample credit runs. It is often the detail that decides whether you move now or wait.

How to lock a price for longer

Three ways: pay a deposit that lets us buy the fabric now, buy the fabric yourself and hold it, or accept a small buffer that covers a reasonable move.

A deposit is the cleanest. Once the fabric is bought and held under your order, the biggest line is fixed.

We will not promise a fixed price for six months without buying the material. Nobody can, and a factory that does is planning to renegotiate later.

What happens when a quote expires

When the window passes, we re-check the fabric and the rate, then send a fresh quote. Often the number is the same or close. Sometimes it is not, and we show you exactly which line moved.

We never silently raise a price. If the new number is higher, you see the old line and the new one side by side.

If the new number is lower, you get that too. It has happened.

How to read the validity line on a quote

A useful quote states three things: the date it was issued, the date it expires, and which lines are fixed inside that window.

If a quote shows only a total and no dates, treat the number as a rough guide. It may still be right when you order, and it may not, and you will have no way to argue either way.

We write the window on the page. It protects you as much as it protects us.

The re-quote we had to send

In 2022 a client sat on a silk quote for eleven weeks. When they came back, the fabric had risen 14 per cent and the whole order was 380 dollars more.

We sent both versions so they could see it was the fabric, not us. They ordered, and now they ask for a 30-day window and decide inside it.

If you are pricing a first order, the private label program explains how we build a quote from scratch. The case studies page shows real landed numbers, and the FAQ page covers payment and pricing questions.

The production process page shows where the quote sits, before sampling begins.

As a small batch clothing manufacturer we get asked one question before every order: how much fabric does this take? The honest answer depends on the style, the size range and the width of the cloth. Here are the real numbers we use to quote.

A dress is not a fixed amount of cloth. A slip dress in viscose may take 1.6 metres, a lined midi dress can take 2.4. The difference is rarely the size; it is how many panels the pattern has.

We quote consumption per garment in metres, at a stated fabric width. If a quote gives you a number without a width, it is only half a number.

Fabric per garment by category: a working table

These are working averages for a mid-size, at 150 cm width. A slip dress: 1.6 to 1.9 metres. A lined midi dress: 2.1 to 2.4 metres. A blouse: 1.4 to 1.7 metres.

Trousers: 1.3 to 1.6 metres. A lined blazer: 1.8 to 2.1 metres, plus 0.9 to 1.1 metres of lining.

A jumpsuit sits between a dress and a pair of trousers, at 1.9 to 2.3 metres. A short skirt is the lightest item on the list, at 0.8 to 1.1 metres.

Add 10 to 15 per cent for a wider size range. Grading from a size 8 to a size 16 can add half a metre to a dress, because the skirt and bodice panels grow.

These numbers are for planning, not for cutting. The marker decides the final figure.

Why the marker decides your fabric bill

The marker is the layout of every pattern piece on the cloth. How well those pieces nest decides how much fabric a run really uses.

A good marker on a small run reaches 88 to 92 per cent efficiency. A rushed one drops to 80 per cent, and that gap is real money.

On a 30-piece dress run using 2 metres each, 60 metres is needed at 100 per cent. At 90 per cent efficiency you buy 67 metres. At 80 per cent you buy 75.

That 8 metres is the difference between a careful cutter and a hurried one.

Fabric width: 145 cm versus 150 cm

Most of our cloth comes in 145 or 150 cm widths, and a few suitings come at 140 cm. The width decides how many pieces fit across the roll.

A 5 cm difference sounds small until you cannot fit a sleeve beside a bodice. That one piece pushes the whole layout longer.

If you buy your own fabric, send us the exact width before you order. We will tell you if it changes the consumption.

Print matching and nap add five per cent

A plain viscose cuts one way. A printed fabric needs the pattern to line up at the seams, which means extra cloth.

We add 5 per cent for print matching on a small run. For a directional print or a fabric with nap, add 8 to 10 per cent.

Satin is the common trap here. It has a sheen direction, so every panel must be cut the same way, and that costs cloth.

We also allow for the end of the roll. The last metre often carries a flaw, so we plan the cut to avoid it rather than discover it.

How to order the right amount without dead stock

Buy the metres you need plus the marker waste, and no more. Extra cloth looks like safety, but on a 30-piece run it usually becomes dead stock.

If a fabric carries a 300-metre dye minimum, a 30-piece order cannot reach it. Use a stock fabric instead and keep your money.

Ask what the leftover will be, and where it will be stored. We hold a client's balance for 12 months at no charge.

What we do when a run comes up short

If the cloth runs short mid-cut, we stop and tell you the same day. The options are to buy another roll, to switch a colourway, or to cut fewer pieces.

We do not cut a short panel and hope. A short panel becomes a remake, and a remake costs more than the metre it saved.

A new roll can also bring a new dye lot, so we check the shade against the first roll before we cut it.

The time we bought exactly enough

In 2021 a client ordered 120 metres for a 60-piece run, calculated to the metre. The marker came in at 118 metres and we had two metres spare.

Then one roll arrived with a two-metre flaw. We had no spare in the same dye lot and had to re-buy 30 metres for four garments.

Now we build a small buffer into every quote and tell the client what it is for. Buying to the exact metre saves nothing when one flaw stops the line.

How to ask for a consumption figure

Ask three things: metres per garment, the width it assumes, and whether marker waste is included. A clear answer names all three.

If a factory will not give a consumption figure before you pay a deposit, treat that as a warning. It is a normal part of a quote.

You can compare two factories on this line alone, and the gap often explains the price difference.

If you are planning a first order, our small batch program shows how we quote fabric and waste together. The case studies page has real landed numbers, and the FAQ page answers the sourcing questions we get most. The production process page shows where fabric is booked, before sampling begins.

A high end clothing manufacturer is judged on the small parts as much as the big ones. A zipper that catches or a button that chips will end a garment's life long before the cloth wears out. Here is what we specify, and why.

Hardware is a small share of your cost and a large share of your returns. A dress with a broken zip comes back; a dress with a plain zip does not.

We put every zip, button, hook and eye on a trim card before bulk, so there is one approved version to match against.

Zippers: nylon, metal and the pull test

For dresses and blouses we use nylon coil zips, usually a size 3. For trousers and jumpsuits we use a metal or a heavier nylon zip, size 5.

The brand matters less than the pull test. We hang 6 kg from a closed zip and check it does not split or shift.

A cheap zip passes on the bench and fails after fifty wears. The difference is in the teeth and the slider, not the tape.

We also check the zip length against the finished opening. A zip cut 1 cm short will not close at the top, and it is a remake, not a repair.

Buttons: corozo, horn and the plastic that chips

Corozo and horn are the finishes clients ask for on a high-end garment. Both are natural, both take dye, and both cost more than polyester.

Polyester buttons are fine on a lined blazer or a casual blouse. They chip at the edge when a garment is pressed too hot, so we lower the iron.

We test buttons by dropping a fixed weight on them and by washing a sample at 30 degrees. A button that cracks in the wash is a refund.

Hardware finish: why plating peels

A gold or gunmetal finish is a plating over a base metal. If the base is thin or the plating is rushed, it peels at the edges after a few wears.

We ask the supplier for the plating thickness and we salt-spray a sample. A finish that survives 24 hours is acceptable for most garments.

Matching finishes across a zip, a button and an eyelet is harder than it sounds. Different suppliers plate differently, so we approve them together.

Nickel release and the EU rule you cannot skip

Nickel can cause skin reactions, and the EU limits how much a metal part may release. This applies to zips, buttons, snaps and rivets that touch skin.

We ask for a nickel release statement with the test method named. A supplier who cannot name the method has not tested.

If you sell in the EU, put this on your trim list before you order, not after.

Custom hardware and the 1000-piece minimum

Custom hardware means a logo zip pull, a branded button or a debossed snap. The tooling is made once and the parts are stamped from it.

Most suppliers set a 1000-piece minimum per design because of the tooling cost. On a 30-piece run that is more parts than you will use for years.

For a first order, use a stock hardware and add your identity on the label. Save custom hardware for a style you know will repeat.

What to write on your trim card

A trim card lists every part with its supplier code, colour, size and finish. It also holds the approved sample of each part.

We keep one card per style and match every bulk shipment against it. If a part changes, the card changes and you approve it again.

A trim card is what stops a factory from substituting a cheaper zip halfway through a run.

The hardware failure we shipped once

In 2020 we shipped a run of jumpsuits with a metal zip we had not salt-sprayed. Two weeks after delivery the client sent photos of peeling sliders.

We replaced the zips on the affected garments and changed our trim process to test every new metal part. It cost us the margin on that order.

We would rather test a 50-cent part than explain a 200-dollar return.

Hooks, eyes and the parts nobody checks

A hook and eye is a two-cent part and a common failure point. We test the tension so a waistband does not pop open on a wearer.

Snap fasteners are another small risk. A snap pressed too hard cracks the cap, and it fails weeks later, usually after the first wash.

Rivets on a jumpsuit pocket are decorative on many designs. If they touch skin, they fall under the same nickel rule as a zip.

We list every one of these on the trim card, even the ones you will never see. A hidden part still carries your label.

How to test hardware before you approve

Ask for three things: the pull test result, a wash test at the care temperature, and a plating or nickel statement. Any supplier should have them.

If they cannot provide them, ask how they test instead. A clear answer is fine; a blank one is not.

You can also wear the sample for a week. A zip that still runs smoothly after seven days is a good sign.

If the style has a hidden zip, ask the factory to test it in the finished sample, not on a loose part. A zip behaves differently once it is set into a seam.

Our high end production page explains how we approve trims before bulk. The case studies page shows real runs, and the FAQ page covers trim questions. The process page shows where trim approval sits in sampling.

Before a cut and sew manufacturer cuts anything, the colour has to be approved. Clients often send a photo and expect that exact navy, then are surprised when the cloth arrives a shade off. Here is how colour approval actually works.

Colour is one of the few things a small brand cannot fully control from a screen. A phone photo, a monitor and a fabric all read colour differently.

We approve colour in writing before the fabric is bought. That single step prevents most of the arguments we used to have.

What a lab dip is and how many you get

A lab dip is a small swatch of fabric dyed to your target colour. The mill makes it so you can approve a shade before it dyes the whole roll.

We usually ask for two or three dips per colour, spaced slightly around your target. You pick one, or ask for one more.

For a stock fabric there is no dip, because the colour already exists. You approve against a physical swatch cut from the roll instead.

Why your navy and our navy differ: Delta E

Delta E is a number for how far two colours sit apart. Under 1 is a close match, 1 to 2 is visible to a trained eye, and over 2 is a different colour.

We aim for a Delta E under 1 on a solid shade and under 1.5 where a print is involved. Anything above that we flag before cutting.

This is the number to ask for. It turns a subjective argument about navy into a measurement.

Light sources: office light versus daylight

A colour looks different under a warm bulb, a cool bulb and daylight. This is why two people can approve the same swatch and disagree.

We check a dip under daylight and under a standard light box. If it shifts badly under one light, we say so before you approve.

If you approve from a photo, tell us the light you used. A photo taken at night under a lamp can look warmer than the cloth.

Matching trim to the fabric: buttons, zips and thread

Trim should match the fabric, not the photo. A zip or a button dyed to its own recipe can sit a shade off the cloth it sits on.

We ask the supplier to dye trims to the approved fabric swatch. For thread we match to the fabric to hide the stitch, and go to a contrast only when the design calls for it.

Buttons are the hardest. A natural button takes dye unevenly, so we approve the batch, not a single piece.

Dye lots and the second roll that does not match

A dye lot is one run of dyeing. Two lots of the same recipe can still differ, and the second lot is where small orders get caught.

We ask for the lot number on every roll and keep it on the cut ticket. If a run needs two lots, we tell you and we check the join.

On a small order it is often better to buy all the cloth from one lot, even if it means buying a little more.

We also photograph each roll under the light box and send the images with the lot numbers, so you have a record of what was cut.

How to approve a colour when you are not in the room

Ask us to mail the physical dip. A photo is a first look, not an approval.

Sign off in writing with the dip number and the date. That record is what we cut against.

If you cannot receive a dip in time, we can hold the colour and send photos under two lights, but we will note that the approval is provisional.

We keep the approval sheet with the dip, the lot numbers and the date, so a question two years later still has an answer.

The colour we got wrong in 2021

In 2021 a client approved a navy from a phone photo taken under a warm lamp. The dip matched the photo, but the photo did not match the cloth.

When the bulk arrived she saw a cooler navy and we had to re-dye twelve metres for the trim. It cost four days and a re-cut.

Now we send a physical dip for any colour the client has not held in hand.

Pantone numbers and why we ask for a physical reference

A Pantone number is a useful starting point, but it is printed on paper, not cloth. The same number looks different on satin and on viscose.

We ask for a Pantone as a direction, then match the dip to a reference you can hold. The number narrows the search; the swatch settles it.

If you only have a Pantone, we will still make dips and show you where the cloth lands. That gap is normal, and it is worth seeing early.

What to write on the colour approval sheet

A good sheet records the colour name, the dip number, the Delta E, the light source used and the date of approval. It also names who approved it.

We keep the approved dip in the sample room beside the approved garment. It is the reference for every future run.

If you reorder in a year, that sheet is why the colour comes back the same.

Our cut and sew service page shows where colour approval sits in the process. The case studies page has real runs, and the FAQ page covers fabric and trim questions. The production process page maps the steps before cutting.

When a client asks why a cut and sew manufacturer charges what it charges, the honest answer is minutes. Every style has a sewing time, and that time is a line on your quote even when it isn't printed. Two factories can quote the same dress eighteen dollars apart and neither one is lying.

So what is SMV, and why does it decide your price? SMV stands for standard minute value. It is the time a trained operator needs to finish one garment at a normal pace, counted in minutes. Multiply it by the factory's minute rate and you have the sewing portion of your quote before you open the file.

Here is how we use it, where it goes wrong, and the question to ask on the next quote you receive.

What a standard minute value actually measures

SMV covers the sewing room only. Joining panels, setting sleeves, inserting zips, hemming, and the small finishing the operator does at the machine. It does not include cutting, pressing, inspection or packing, which sit on separate lines.

On a 30-piece run those extra stages add roughly 20 to 35% on top of the sewing minutes. If a quote shows a sewing number and nothing else, ask what happened to pressing and inspection.

Our own ranges, for reference: a slip dress 20 to 30 minutes, a lined midi dress 40 to 60, trousers 35 to 55, a lined blazer 90 to 150. Structured garments cost what they cost because of minutes, not because of a brand name.

Why the same style carries two SMV numbers

A textbook SMV assumes a steady line. A 30-piece run is not a steady line. The first ten pieces of any style are slower while the operators learn the seams, and the last ten are the fastest.

Line efficiency on a small run usually lands between 60 and 75%. On a long bulk run it reaches 85 to 90%. Take a 45-minute SMV: at 85% efficiency that is 53 actual minutes, and at 65% it is 69. Sixteen minutes, same garment, same operator.

This is why we quote small orders on effective minutes rather than textbook SMV, and we write both numbers down.

How minutes turn into money: the minute rate

Our small-run minute rate sits around RMB 1.1 to 1.5 a minute, roughly 15 to 21 US cents. It covers the operator, the line supervisor, machine time and floor overhead, and nothing else.

Run the numbers on a lined midi dress. Fifty SMV, 65% efficiency, 77 actual minutes, and at RMB 1.3 that is about RMB 100, or $14 of sewing. Fabric, trims and freight are separate lines.

The same dress at 1,000 pieces: 85% efficiency, 59 minutes, and a minute rate near RMB 0.9 because the line never stops learning. That comes to about RMB 53, or $7.40. The unit price falls with quantity because of arithmetic, not because of a volume discount.

How to compare two garment quotes using SMV

Ask two questions: what is the SMV, and what is the minute rate. If a factory says it doesn't use SMV, ask for minutes per garment by operation. Anyone who quotes sewing can produce that.

Then sanity-check it. A lined blazer with a chest canvas cannot be sewn in twenty minutes. If a quote implies that, the price is either missing an operation or the lining is fused and minimal. Both are fine, but you should know which one you bought.

A quote that lists sewing as one lump sum, with no minutes and no operations, tells you nothing about where the cost sits. That is the one to push back on.

Where SMV does not help you

SMV tells you the sewing cost and little else. It will not tell you whether the fabric is right, whether the grade holds at size 14, or whether the same team will still be there for your reorder.

It is also not a number you can compare across countries without the minute rate. A 60-minute garment in Guangzhou and a 60-minute garment in Portugal are the same time and very different money.

Use it for one job: checking that a quote was built from operations rather than from a guess.

The run where our SMV was wrong

In 2022 we quoted a lined jumpsuit with an invisible zip running through the waist seam at 55 minutes. The first pieces took 78. Thirty pieces at 23 extra minutes and RMB 1.3 is about RMB 900 that we absorbed rather than re-quote.

We now cut and time a first-piece sample before quoting anything with a hidden closure, a bias cut or a lined bodice. It costs us half a day and saves the argument.

Minutes are one of several lines behind a price. Our cut and sew service page sets out what the sewing price includes. The case studies page shows real runs with their timelines, the FAQ page answers costing questions, and the production process page maps the steps from tech pack to carton.

How a private label clothing manufacturer packs your order decides what your first stockist sees. A garment can be perfectly sewn and still arrive looking tired, because packing is where small brands quietly lose presentation points. The rules are not ours; most of them come from retailers and customs.

So what does packing actually involve on a small order? A polybag, a hangtag, a fold, a carton and a label. Five items, and each one has a rule attached that you can check before you approve.

Here is what we do, what it costs, and the packing mistake that cost us money at destination.

Folding, pressing and polybags: why garments arrive creased

Every piece is pressed on a board and folded to a fixed size before it is bagged. The fold size matters. A stockist with standard shelving wants a consistent rectangle, not whatever shape the garment falls into.

We use 30 micron LDPE as standard, with recycled LDPE available on request. A bag with an opening of 12.7cm or more needs a suffocation warning printed on it for the US market, and we print it as standard rather than waiting to be asked.

Creasing is where fine fabric punishes you. Satin and viscose mark along a fold and hold the mark. We tissue-paper the fold line on both, and leave polyester and TR suiting without it because they recover on their own.

In 2021 we shipped forty viscose dresses flat with no tissue. They arrived with visible lines across the bust and the retailer asked for a re-press. That cost RMB 12 a piece at destination, more than the tissue would have cost for the whole run.

Hangtags, barcodes and care labels: the retail rules

Hangtag attachment is a small decision with a visible result. A plastic barb is fast and cheap, but on fine knits it can pull a hole. String loops take longer and sit better on silk and light viscose.

Your barcode is yours, not ours. A GTIN comes from GS1, costs roughly $250 to set up, and stays with your brand. We print it on the tag, about 1cm from the edge, on the back.

Care labels are a legal line, not a design line. Fibre composition and the care symbols have to be in the language of the market you sell into, which in the EU means the member states you actually ship to. We ask for that list before the tags go to print.

On a 30-piece run, 500 is often the smallest quantity a tag supplier will print. The leftovers are normal, and it is the same pattern that catches clients on woven labels and custom bags.

What packing costs per piece on a small run

Packing is small money per piece, and it adds up. Our typical 30-piece breakdown: polybag and sticker $0.06 to $0.12, printed hangtag $0.15 to $0.40, attach labour $0.05 to $0.10, tissue and folding $0.03 to $0.06, carton and tape about $0.10.

That is roughly $0.40 to $0.80 a piece, and it is worth budgeting before you fall in love with a box. A rigid luxury box with ribbon adds $1.50 to $4.00 per piece, plus volume in the carton, which air freight charges you for twice.

The tag is usually the largest line, because it carries a print minimum. One tag design used across a whole season is the cheapest way to hold that number down.

Cartons, weights and how thirty pieces leave the building

A standard 60 by 40 by 40cm carton holds about fifteen dresses or eight lined blazers. We keep gross weight under 23kg so one person can lift it without a trolley.

Thirty dresses is normally two cartons. Each one gets a label with the SKU, colour, size ratio, quantity, gross and net weight, and a carton number, so carton 1 of 2 can be identified without opening it.

Air freight bills on volume weight, calculated as length times width times height in centimetres divided by 5000. A half-empty carton costs the same as a full one, which is why we consolidate before sealing rather than after.

The packing list and the photo that ends arguments

Every carton ships with a packing list: SKU, sizes, quantities, weights and dimensions. We photograph the open carton before it is sealed, and that photo goes into the file beside the packing list.

It is the reason a claim about a missing size or a wrong label gets settled in one email instead of two weeks. We started doing it in 2019, after a client found four cartons with mismatched labels and neither side could say where it had happened.

Packing sits at the end of a run, but it is decided at the start. Our private label service page covers what we handle, the case studies page shows finished orders, the FAQ page answers labelling questions, and the production process page shows where packing falls in the schedule.

A small batch clothing manufacturer will normally ask for 30% with the order and the remaining 70% before the goods leave. That split is standard and not worth arguing about. What matters far more is what the 70% is tied to, because that is the part that protects you.

So when is the balance actually due? Not on a date. On a set of documents that prove the order is finished and inspected. If your terms say balance before shipment, ask for the list of documents that triggers it, in writing.

Here is what we send, when we ask, and the order where we asked too early.

The balance is a document trigger, not a date

Four documents do the work: the pre-shipment inspection report, the packing list, the commercial invoice, and the transport document, which is a bill of lading copy for sea or an air waybill for air.

Why does this matter? If the balance is tied to a date, the date arrives whether the goods are ready or not, and you pay on a promise. If it is tied to documents, you pay on evidence. The same 70% behaves very differently under those two sets of terms.

We write those four documents into the proforma invoice, together with the incoterm, so there is no version of the story where the balance is due before the report exists.

What we send before we ask for the balance

The pre-shipment report comes first, with the defect counts and the AQL level used. Then the measurement sheet per size against the agreed tolerances, the carton photos, and the packing list with weights and dimensions.

The commercial invoice follows, matching the proforma exactly: same description, same HS code, same incoterm, same currency. Small mismatches here cause customs delays far more often than the duty rate does.

For clients who cannot visit, we add a short video of the cartons being sealed. It takes two minutes and removes most of the anxiety around paying a factory eight time zones away.

When the inspection fails: holdback, rework, re-inspect

Minor defects inside the agreed AQL: we ship, and the count sits on the report so you know what to expect. Major defects above the threshold: we rework or remake before we ask for anything.

Four garments needing a remake takes five to seven days. We absorb that rather than pushing the balance request and letting the problem travel to you.

Some brands ask to hold 10% for thirty days after arrival on a first order. We have agreed to that twice. We do not agree to it on a reorder, where the sample is approved and the risk is known.

What we never accept is the reverse: a balance paid before any inspection exists. On a first order, that is the line to hold.

Red flags around the final payment

The balance is demanded before an inspection report exists. The account name on the invoice does not match the company name. Payment is pushed toward cash or a personal account, which no legitimate factory needs.

The one to watch hardest is a last-minute change of bank details by email. That message pattern is fraud almost every time it appears, and it does not come from the factory. Call the person you have been dealing with, on a number you already have, before any wire moves.

And if there is no packing list, no carton numbers and no measurement sheet, the order is not finished, whatever the calendar says.

The 2020 order where we asked too early

In 2020 we sent a balance request with a photo of finished garments on the table, before the pre-shipment inspection had run. The client paid within a day. Four cartons then turned up with the wrong care labels.

We reprinted thirty labels and air-freighted them at our own cost. It was a small sum and a fair hit, because we had asked for money against a photo.

Since then the balance request goes out with the inspection report attached and nothing else. Clients pay faster, not slower, when the evidence arrives with the ask.

Reorders: when the balance can move earlier

On a second run of the same style, with an approved sample on file and the same fabric lot, we will sometimes accept 50% on order and 50% after arrival. That is not a discount; it is a lower risk profile.

Ask for it after a clean first run, not before. A second order is cheaper to set up anyway, which is exactly why the risk on it is smaller.

Payment terms sit next to everything else in an order. Our small batch production page covers what we run and at what size, the case studies page shows completed orders, the FAQ page answers payment and shipping questions, and the production process page maps the stages that produce those four documents.

As a high end clothing manufacturer, the question we hear most after a sample lands is why the bulk does not feel the same. The short answer is that your mock-up was probably cut from a different roll than the bulk will be. That gap is normal in this trade, and it is the biggest single reason a first run disappoints. Here is what actually changes between sample and bulk, and how to close the gap before you approve.

Do the sample and the bulk have to be identical? Ideally yes, and in practice they match on specification rather than on roll. A 30-piece order needs 60 to 110 metres, which is a different purchase from the two or three metres a sample room works with. Ask your factory which of the two you are looking at, because the answer tells you how much risk sits inside the order.

Sample yardage and bulk yardage rarely come from one lot

Sample yardage is bought in short lengths, often two to five metres, from whatever the mill has on the shelf that week. Bulk yardage is a mill order in its own right, with its own dye lot number printed on the roll.

Even when composition and weight match on paper, a mill can finish a five-metre length and a three-hundred-metre length on different days. The finishing bath, the drying tension and the calendar pressure each move the hand of the cloth a little.

So a sample can feel softer and a bulk roll flatter while the spec sheet reads identically on both. That is not a false document. It is a difference the document was never designed to catch.

Why a mock-up can read lighter or heavier on a hanger

Weight is quoted in grams per square metre, and the tolerance we accept is plus or minus five per cent. On a 120 gsm viscose that is a 12 gsm spread, which is enough to change how a skirt falls.

Colour is the other quiet mover. Two rolls from the same nominal lot can sit one to two Delta E apart, and under a shop light that shows up across a rail of ten garments.

A third factor has nothing to do with the mill at all: time. A bias-cut satin panel relaxes after it hangs for two days. If the sample was measured on day one and the bulk on day three, the numbers will disagree even when the cloth is identical.

Three questions to ask before you approve a sample

Was this sample cut from the roll reserved for bulk, or from stand-in yardage? If it was stand-in, what are the composition, weight and finish of the bulk roll, in writing? Can you keep one metre of the sample roll as a reference card?

Those three answers take a factory ten minutes and can save a brand a month. A reference card is a physical record you can lay beside the bulk roll and compare with your own eyes.

If nobody can answer the first question, treat that as the signal. It usually means the sample and the bulk are being sourced by two people who have never compared notes.

What we do so the mock-up and the bulk match

On every style we try to buy the sample yardage from the same mill order as the bulk. That works when the brand uses stock fabric and the mill holds enough of one lot. It does not work when the fabric is dyed to order, because a 300-metre minimum arrives as a single lot long after the sample was approved.

When a sample has to be cut from stand-in yardage, we write that into the spec sheet in plain words and send a photo of both swatches side by side. The client signs off on the difference before we cut, not after.

We also keep the first metre of every bulk roll as a shade band. If a second roll turns up darker, the cutting room works from one band at a time instead of mixing rolls across a bundle.

None of this removes the gap between a sample and a first run. It only moves the surprise to a moment when it costs nothing.

What a fabric card must prove about quality

High quality is not a claim, it is a set of numbers you can check. We print four on every fabric card: composition, weight in gsm with a tolerance, shrinkage after a 30-degree wash, and a colourfastness grade for wash and rub.

Those four numbers are what turn a mock-up from a photograph into a specification. Without them a sample is a nice object and the bulk is a guess.

The 2021 order where two lots met on the rail

In 2021 we made a 40-piece viscose dress in navy. The sample came from a stock roll, and the bulk came from a fresh lot that read half a shade deeper. We checked both internally, found them inside the mill's own tolerance, and did not flag the difference.

Twelve of the forty dresses were cut from the second roll, and on the client's rail the gap was plain to see. We re-cut twelve panels from a single lot and lost nine days on a run that was already tight.

Since then the rule is fixed. If two lots are going into one order, the client sees both before we cut, and the decision goes into the file.

Where fabric sits in the rest of your order

Fabric is one variable in a run, and the same logic covers colour approval, trims and the sample itself. Our high end production page lists the standards we hold on every style, the case studies page shows runs where this went right and wrong, the FAQ page answers fabric and sampling questions, and the production process page maps where fabric approval sits in the schedule.

Every cut and sew manufacturer has a rework rate, and most will not tell you what it is. Ours runs three to eight per cent on a first run and one to three per cent on a reorder. Those numbers sound small until you price what a single remake actually costs at thirty pieces. Here is the honest arithmetic, and the four things that keep the rate down.

What is a rework rate? It is the share of pieces that fail a check and have to be fixed or remade before shipping. It is not the same as a defect rate, because a defect can sit inside tolerance while a rework always costs somebody money. Ask which number you are being shown.

Rework and remake are not the same thing

Rework means the original garment is fixed. A seam re-stitched, a button replaced, a hem re-pressed, a zipper swapped. The piece stays in the run and the cost is minutes.

A remake means the piece is cut again from new fabric. The original is scrapped, and you pay for cloth, trims, cutting and sewing a second time. A remake always costs more than the garment's own FOB.

Keeping those two words apart matters, because a factory that reports one number for both is hiding the expensive half.

What a single remake costs on a 30-piece run

Take a lined viscose dress at a $72 FOB. A remake consumes about 2.1 metres of fabric at $4.20, which is $8.80, plus lining at $1.60, plus trims at $2.10, plus cutting and sewing time at roughly $14. That is $26.50 before shipping, or 37 per cent of the piece's own value.

At 30 pieces, a five per cent remake rate is one and a half garments, so call it $40 in direct cost. That looks survivable until you add the days: a remake batch takes five to seven working days and lands in the middle of the shipping window.

So the real cost of rework is not the money. It is the week.

Why small runs carry a higher rework rate than bulk

A 30-piece order is a learning curve. Operators have never sewn this style before, and the first ten pieces teach the line where the tricky steps are. On a 500-piece order those ten pieces are two per cent of the run. On thirty they are a third of it.

The second reason is bundling. Small runs move in small bundles, so one operator handles several operations on the same garment. That flexibility is why we can take 30 pieces at all, and it is also why a mistake can travel further before anyone catches it.

The third is fabric behaviour. A new roll that shrinks differently or presses differently will show up in the first pieces, not the last.

Four checks that hold a rework rate down

First, a sealed pre-production sample with the spec sheet and the trim card beside it, so the line works from a physical reference rather than a photo. Second, in-line checks at the end of each operation, not only at the end of the line.

Third, a first-off piece from each size approved before the rest of that size is sewn. Fourth, a measurement table that names the five critical points and the tolerances at each one, so a check is a number rather than an opinion.

None of those four is expensive. They are all scheduling decisions made before cutting starts.

One number is worth tracking over time. If the same style comes back on a second run with the same fault, more checking is not the fix. The fix is a change to the sample, the spec sheet or a machine setting, and it goes into the file before the reorder starts.

Who pays for a remake, and when it is not our fault

If the fault is ours, meaning a sewing, cutting or trim error against the approved sample, we absorb it. That is not goodwill. It is the cost of our own mistake.

If the fault is a design change made after approval, the client pays, because the sample on file is the reference and it moved. If the fault is fabric the client supplied that behaves outside the stated spec, we flag it before cutting rather than after.

The one case worth arguing about is a tolerance nobody wrote down. In 2019 we had no written hem tolerance, so a two-centimetre difference became a negotiation instead of a check. We now write tolerances into the spec sheet for every style, and a rework is decided by a number.

The 2021 run where eight blazers went back on the line

In 2021 we ran 30 lined blazers. The sleeve pitch was slightly off, which only shows when the jacket is on a body rather than a hanger. Eight pieces had to have their sleeves re-set.

Re-setting took six working days and pushed the shipment back by a week. We caught it at the end-of-line check rather than at pre-shipment, which is the only reason it cost a week and not a month.

After that run we added a wear test on the first-off piece. The garment goes on a mannequin and gets worn for an hour before the size is released. It is a slow start and a fast finish.

Where rework sits in a small run

Rework is one number in a run, and it connects to sampling, cutting and inspection. Our cut and sew service page shows what we take on and what you supply, the case studies page shows finished runs, the FAQ page answers quality and timing questions, and the production process page maps the three checks that catch a fault before it ships.

As a small batch clothing manufacturer, we get asked for thirty pieces in four colours almost every week, and the honest answer is that thirty pieces is a quantity, not a colour plan. One style in four colours is four small production problems wearing one order number. Here is what the minimum actually applies to, and how to split thirty pieces without multiplying your cost.

So how many colours fit in thirty pieces? With stock fabric, two or three, at ten to fifteen pieces each. With fabric dyed to order, one, because the dye minimum is usually 300 metres and thirty pieces need about 70. That single number decides most of the answer.

Per-style, per-colour and per-fabric are three different minimums

A factory quotes one minimum, and it is almost always the per-style figure, which for us is 30 pieces. That is the total across all colours and sizes, not a number for each colourway.

The per-colour minimum is a different thing and it is rarely written on a quote. It exists because each colour needs its own cutting lay, its own bundle tickets and its own quality record.

The per-fabric minimum is the third and the hardest. If two colours come from two different fabrics, the order is really two orders sharing one due date.

Stock fabric lets you split; dyed-to-order does not

Stock fabric is bought by the metre from a mill's existing range, so ten metres of ecru and twenty of black is a normal purchase. That is what makes two or three colours possible at thirty pieces.

Dyed-to-order fabric is made for you, and the mill will not start the machine below its minimum, which runs 300 metres on most viscose and 500 on prints. Thirty pieces of a lined dress need about 70 metres, so a single colour already leaves roughly 230 metres on the shelf.

Ask which type your fabric is before you plan colours. It is the difference between a two-colour capsule and a one-colour run.

What each extra colour actually costs

A lab dip costs $35 to $90 per colour, and most mills give two or three dips around the target. Three colours means three approval rounds, each of which takes two to four days.

Each colour also needs its own cutting lay and its own marker. One marker for a five-size nest is fine. Three colours of fifteen pieces each cannot share it, so you pay the laying time three times.

Trims behave differently. Woven labels, care labels and hangtags are ordered per design rather than per colour, so three colours do not triple the trim order. Only custom-coloured trims, like a dyed zipper tape, follow the fabric.

A worked example: 30 pieces, two colours, one style

Say a midi dress in viscose, thirty pieces, split 15 black and 15 ecru, both from stock fabric. Fabric is 2.1 metres per piece plus eight per cent for the marker, so about 34 metres per colour.

Stock viscose at $4.20 a metre makes the fabric $143 per colour. Two lab dips at $60 each add $120, and the second marker adds around $60 of laying time. The extra colour lands at roughly $180 on a $2,160 order.

Six per cent for a second colourway is a fair trade if it doubles the ways a customer can buy the dress. Four colours at that rate is not a capsule, it is a wholesale range.

The 2021 order with four colours and eight pieces each

In 2021 a client asked for four colours, eight pieces each. We took it because the total was thirty-two pieces and the fabric was stock. What we did not price properly was the cutting room.

Four colours meant four lays, four sets of bundle tickets and four quality records for a run that sewed in a day. The laying and ticketing took two extra days and cost us about $310 that we had not quoted.

We finished the order and lost money on it. Since then a third colour gets its own line on the quote, so the client can see what it costs and decide whether it is worth it.

How to plan colours at thirty pieces

Two colours, fifteen pieces each, on stock fabric, is the sweet spot. Three colours works if you accept eight to ten pieces each and the fabric is stock. Four or more only makes sense if you are planning a reorder before the first run ships.

If the fabric must be dyed to order, plan one colour for the first run and add a second on the reorder, when the dye minimum is already paid and the pattern is proven.

One more option is worth knowing. A single colour in two finishes, such as a matte and a satin version of the same base, reads as two styles on a rail and stays inside one dye minimum.

Where colour planning sits in the order

Colour count affects fabric buying, cutting and pricing, which is why it belongs in the first conversation. Our small batch production page sets out what we run at what size, the case studies page shows completed runs, the FAQ page answers MOQ and colour questions, and the production process page maps where fabric approval sits in the schedule.

Our Fashion Pains Solver tool, still in training, will let a brand test colour and minimum combinations before they ever email us. Until it is online, a two-line message about colour count and fabric type gets you the same answer.

Bubbles in a fused front panel come from one of three things: heat, pressure, or a shrinkage mismatch between the shell and the interfacing. As a high end clothing manufacturer, we test the bond before a structured garment is cut, because a bubble does not show on the fitting table. It shows after two washes, in your customer's hands.

So what actually causes it? The interfacing is a base cloth coated with heat-activated resin dots, and when those dots melt they have to flow into the weave of the shell fabric and set. If the shell then moves differently from the interfacing, whether in a wash or under a steam iron, the two layers separate in small patches. That patch is the bubble.

There is no way to repair a bubble after the garment is made. It has to be prevented at the fusing stage, and that stage happens before a single seam is sewn.

Fusible interfacing shrinkage: the mismatch that starts the bubble

Every fabric moves when it gets wet. Viscose can shrink four to six per cent, linen two to three, polyester under one. A fusible base cloth is usually woven or knitted to be stable, so its figure sits closer to one per cent.

Put a four per cent shell over a one per cent interfacing and the shell contracts while the glue layer does not. The difference has nowhere to go, so it puckers. That is why we bond a test square and wash it before a structured order is cut.

The test takes ten minutes. Bond a ten by ten centimetre square at production settings, wash it at thirty degrees, dry it flat, then look at the corner lift and try to peel the layers apart by hand. If they separate, the pairing is wrong.

Heat, pressure and dwell time: three settings a press operator owns

A fusing press has three dials and all three matter. Polyamide-coated fusibles on viscose and TR suiting generally want around 150 degrees Celsius, three to four bar of pressure, and ten to fifteen seconds of dwell.

Too hot and the resin strikes through to the face of the fabric, which reads as a shiny patch under raking light. Too cool and the dots never fully flow, so the bond is weak from day one and fails on the third wear.

Too little pressure and the fabric lifts before the resin sets. Too much and the weave is flattened, which is visible on a satin or a fine suiting blend. We keep a fusing log per style, so the settings are a record rather than somebody's memory.

One setting that gets forgotten is moisture. Damp shell fabric traps steam under the resin, and trapped steam is the most common cause of a blister that appears three weeks later.

A second press, sometimes called curing, is worth the extra minute. It re-heats the bonded panel without pressure, which lets any trapped moisture escape and lets the resin finish setting. Panels that travel straight from the fusing machine to the sewing line are the ones that come back.

What to ask before a structured jacket is cut

Ask for the interfacing specification in writing: base cloth, resin type, weight in grams, and the recommended fusing temperature. A quote that says fused front tells you nothing about whether the piece survives a season.

Ask whether the interfacing was tested against your shell fabric, not against a standard swatch in a mill catalogue. The two have to be a pair, and the pair has to be washed together.

Ask what happens if the bond fails. We cover a failed bond as a remake, because it is our setting or our pairing, not your design decision.

Fusible or sew-in canvas: the tailoring choice behind the price

On a lined blazer you have two ways to build the front. A fusible is glued in by machine and takes minutes. A floating canvas is attached by hand and takes twenty-five to forty minutes per jacket.

At a Guangzhou minute rate of roughly fifteen to twenty-one US cents, that hand work adds four to eight dollars per jacket. On a thirty-piece run it is one hundred and twenty to two hundred and forty dollars across the order.

A canvas front rolls with the body and recovers from pressing. A fusible is flatter, faster and cheaper. Neither is wrong, but the difference is a decision worth making before sampling rather than after.

The lapel run where twelve fronts had to be re-cut

In 2022 we ran thirty lined blazers for a European label. The shell was a viscose and linen blend. The interfacing was a standard weight we had used many times on TR suiting.

We skipped the washed test square, because we had bonded that interfacing before. We had, but never over that blend ratio. Twelve lapels bubbled at the first steam press, and all twelve fronts had to be re-cut.

Re-cutting cost six working days and about four hundred and eighty US dollars in fabric and time. The fix was a lighter knitted-base interfacing and a fusing temperature about ten degrees lower. That pairing now sits in the style file, and the test square is mandatory on every new blend.

Where interfacing sits in your production schedule

Fusing happens after cutting and before sewing, which means a pairing error is discovered at the worst possible moment. Our high end production page sets out the standards we hold, the case studies page shows finished runs, the FAQ page answers construction questions, and the production process page maps where sampling and approval sit ahead of bulk.

Off-grain cutting and a weft that is not square to the warp are the two reasons a skirt hem spirals after washing. Every cut and sew manufacturer has a number for this, and ours is two per cent skew on plain wovens. Above that we ask the mill to replace the roll.

How do you know a roll is skewed before it is cut? You measure it on the table. Lay a metre of fabric flat, square one corner against the selvedge, and see how far the weft yarn drifts along the opposite edge. A two centimetre drift over one metre is two per cent skew, and that is the ceiling we work to.

This is a fabric property, not a sewing fault. That is why it has to be checked before the marker is made, not after the garments come off the line.

Warp, weft and true bias: three directions on one roll

The warp runs along the length of the roll, parallel to the selvedge. The weft runs across it, from selvedge to selvedge. True bias sits at forty-five degrees to both.

Woven pattern pieces are cut with the warp following the length of the piece, because the warp is the stable direction. It stretches least, so a seam sewn along it keeps its length through wear and cleaning.

Bias has the opposite behaviour. It stretches most, which is why a bias-cut panel drapes the way it does and also why it grows. A bias panel needs to hang for twenty-four to forty-eight hours before the hem is levelled.

Bias also needs its own marker, because the pieces cannot share a lay with straight-grain pieces without wasting a lot of cloth. Budget ten to twenty per cent more fabric for a bias-cut style, and put the hanging time into the schedule rather than after it.

How fabric skew is measured before a marker is made

Skew is the angle between the weft and a true right angle to the warp, expressed as a percentage. It is tested on a squared sample, either against a mill report or in the workshop.

The thresholds we use are two per cent for plain wovens and one per cent for stripes, checks and anything with a visible line. A stripe that is one per cent skewed will not match at the side seam, however carefully it is cut.

Ask for the number in writing, with the report reference and the date. If the report is more than a year old, ask whether the finishing line has changed since, because a change of stenter settings moves skew.

One practical note from the cutting table. When a roll is slightly skewed but inside tolerance, the marker can be laid so the pattern grain follows the warp rather than the selvedge. That correction costs a little extra fabric and saves the panel, and it belongs on the cut ticket so the next run repeats it.

Why a single-ply yarn twists more after washing

Twisting in the wash starts with the yarn. A single-ply ring-spun yarn carries residual torque from the spinning process, and water releases it. Two-ply yarns balance that torque against each other, which is why they twist less.

The second driver is skew. A panel cut from a skewed roll is already slightly off square, so when the fabric relaxes in water the error grows instead of cancelling out.

The third is the wash itself. A thirty degree home wash with a spin cycle moves fabric more than a cold hand wash, which is why the same garment can behave differently in two customers' hands.

Our test is a fifty by fifty centimetre square cut on grain, washed at thirty degrees and dried flat. If the corners no longer meet, the fabric is not suitable for a garment with a straight hem.

Stripe and check matching: the grain cost nobody quotes

A stripe or a check has to be matched across the seams, which means extra fabric. Matching on the cross-grain adds roughly eight per cent to consumption, and a full check match on a plaid can add fifteen.

On a thirty-piece run needing about seventy metres, fifteen per cent is more than ten extra metres, which is forty to sixty US dollars of fabric plus the extra cutting time.

It also adds cutting minutes, because the pieces cannot be nested tightly. A matched lay is laid one way, and the leftover between pieces is wasted by design. Quote the match before you quote the style.

The 2021 linen that twisted forty skirt panels

In 2021 a client supplied their own linen for a thirty-piece skirt run. The mill report showed skew of four and a half per cent, and we cut it anyway because the delivery date was tight and the client had already paid for the roll.

After the first wash test, the hems spiralled on forty panels across a second, larger order. We re-cut them from a replacement roll, which cost nine working days and a full second cut.

Since then a skew figure above our ceiling stops the order until a replacement roll arrives. It is the one check where we would rather delay a run than start it.

Where grain control sits in your order

Grain is decided at the fabric stage and confirmed at the cutting table, which is why it belongs in the fabric approval step rather than in final inspection. Our cut and sew service page shows what we take on and what you supply, the case studies page shows finished runs, the FAQ page answers fabric and quality questions, and the production process page maps where fabric approval sits in the schedule.

A rush order moves four things and none of the rest. As a small batch clothing manufacturer we can compress quoting, sampling rounds, line scheduling and finishing, but we cannot compress dyeing, fabric relaxation or transit. That is why the honest answer to a four-week request is no.

So how fast can a thirty to fifty piece order actually run? Eight weeks is realistic when the fabric is stock and you approve the first sample quickly. Twelve is normal. Six happens only when every stage lands in the right order and nothing needs a second attempt.

The difference between those numbers is not effort. It is which blocks of the schedule have a fixed length and which ones are waiting on a decision.

The twelve-week timeline, block by block

A standard run breaks into four blocks. Three weeks for quoting and buying fabric, three weeks for two mock-up rounds with approvals between them, four weeks on the machine, and two weeks for finishing, packing and shipping.

Notice that only four of the twelve weeks are sewing. The other eight are preparation and recovery, which is where a rush order is either won or lost.

The clock starts when the deposit lands, not when you first send an enquiry. A week spent deciding is a week off the end of the schedule.

What a small factory can genuinely compress

Quoting is the easiest block to shorten. A clear tech pack with a fabric reference can be priced in twenty-four to forty-eight hours, and we will hold a line slot for five days while you decide.

Sampling compresses if you approve the first mock-up instead of asking for a second. Dropping from two rounds to one saves ten to fourteen days, and it is the single biggest lever available.

Stock fabric removes the dye lead entirely, which is two to three weeks on dyed-to-order cloth. Scheduling can pull a run forward into a gap on the line. Finishing can run a second shift for a week if the volume is small.

Add those up and twelve weeks becomes eight, which is the realistic floor for a small woven order.

What cannot be compressed, whatever the fee

Dyeing has a minimum of about three hundred metres and a lead time of two to three weeks. No payment changes that, because the mill is running a machine, not a favour.

Lab dips take two to four days per round, and most mills want two rounds before the colour is right. Fabric also needs to relax for twenty-four hours before cutting, or the panels move after they are cut.

Sewing minutes are fixed by the garment, not by the deadline. A lined blazer is ninety to a hundred and fifty standard minutes whatever the calendar says. Our three inspections sit inside the production window rather than after it.

Transit is the last fixed block. Air freight is five to seven days and sea freight is twenty-eight to thirty-five. If the garment must be on a rail by a date, the shipping mode is decided backwards from that date.

What a rush order actually adds to your price

Expect ten to twenty per cent on the unit price, and ask which lines it lands on rather than accepting a single figure. On a thirty-piece run with a seventy-two dollar FOB, that is roughly two hundred and twenty to four hundred and thirty dollars across the order.

The main drivers are overtime on the line, a priority slot that displaces another order, and split shipping. Splitting a shipment so half goes by air often costs more than the whole order going by sea.

Expedited trims are the quiet one. Woven labels and custom hangtags have their own two to three week lead, and a rush order does not speed them up.

A second shift is cheaper per garment than air freight, so if the deadline has a few days of slack in it, pay for the shift and keep the sea booking. We price both options side by side so the comparison is visible before you choose.

The 2022 order we promised in nine weeks and shipped late

In 2022 we quoted a nine-week rush on a fifty-piece viscose dress run for a US label. The fabric was stock, the sample was approved in one round, and every block was booked to the day.

We shipped in ten weeks and two days. A trim delivery slipped by four days, and because the quote had no float in it, those four days went straight onto the delivery date.

Since then every rush quote carries one week of float that we do not advertise as float. It looks like a slower promise, and it is the reason our rush dates now hold.

How to plan a rush order without gambling

Send the tech pack before you ask for the date, so the quote is real rather than indicative. Say which fabric you want and whether it is stock or dyed to order, because that one answer moves the schedule by three weeks.

Ask the factory for the two dates that are genuinely fixed and the two that are still open. If nobody can name a fixed date, the timeline is a wish.

Our Fashion Pains Solver tool, still in training, will let a brand test a target date against a realistic timeline before they ever email us. Until it is online, send the tech pack and the fabric type, and we will give you the two dates that matter.

Then decide the shipping mode before production starts, not after. A rush order that saves two weeks in the workshop and loses four at the port is not a rush order. Our small batch production page sets out what we run at what size, the case studies page shows completed runs, the FAQ page answers timing and MOQ questions, and the production process page maps the three stages a rush has to respect.

A cut and sew manufacturer sees two kinds of seam failure, and the fixes pull in opposite directions. One is slippage, where the yarns of the fabric pull apart either side of the stitch line. The other is thread breakage, where the stitch itself gives way and tears the cloth around it. Telling them apart comes first, because the setting that cures one makes the other worse.

Which one is your garment heading for? Four numbers decide it, and all four can be written into a spec sheet before anything is cut: stitches per inch, thread size, needle size and seam allowance. We publish all four on our own specs and price against them. Below is what each number does, the range we hold, and the one test we run before a bulk run starts.

Stitches per inch: the number that controls seam slippage

Stitches per inch, or SPI, is the density of the seam. On a plain woven we run 10 to 12. On fine silk and chiffon we go to 12 to 14, and on heavy wovens down to 8 to 10.

Too few stitches and the load lands on fewer points of contact, so the yarns slide and the seam opens. Too many and the needle perforates the cloth, cutting fibres along the stitch line, which is how a seam tears the fabric instead of holding it.

Slippage is a fabric property more than a sewing fault. Satin charmeuse has long floats and very little interlacing, so its yarns move easily. Loose viscose crepe and open-weave linen behave the same way. Polyester satin slips without breaking, which hides the problem until a customer puts the garment on.

Thread size and needle size: what a fine viscose needs

Thread is measured in tex, and the number rises with thickness. Tex 27 is our general seaming thread for medium wovens. Tex 40 goes on topstitching and denim-weight cloth. Tex 18 to 24 is what a fine silk or a light viscose wants, because a thicker thread drags a thin weave out of line.

The usual rule is that the thread should fail before the fabric does. That way an overloaded seam opens at the stitch rather than ripping a hole in the cloth. On stress seams we break that rule on purpose and match the thread to the fabric instead.

Needle size follows the same logic. We use 70/10 on silk, 80/12 on viscose and light cotton, 90/14 on medium wovens and 100/16 on anything heavy. A needle that is too large leaves a visible perforation line down the seam allowance.

Seam allowance at stress points: why 1cm is not always enough

One centimetre is our standard seam allowance, and it is fine across most of a garment. At four places we widen it to 1.5cm: the side seam at the hip, the armhole, the crotch, and the centre back on a structured piece.

The reason is movement. A seam with more allowance has more cloth to absorb a pull before the yarns at the stitch line start to slide. On a fitted dress graded tight at the hip, 6mm of allowance is not a seam, it is a stress test.

Stitch type matters here too. A lockstitch 301 is the least extensible and the most common. A chainstitch 401 stretches roughly 30 per cent more, which is why it belongs on a knit or an inseam. A five-thread overlock 516 on a side seam is strong but adds bulk, and on a sheer fabric you can see it through the cloth.

How we test a seam before the bulk run starts

We cut a seam sample from the actual production roll, sew it on the same machine with the same thread and needle, then pull it to a set load and measure the opening. On plain wovens we stop at 6mm. On satin and silk our ceiling is 4mm.

The second number is seam efficiency, which is the strength of the seam as a percentage of the strength of the fabric. We aim for 80 to 90 per cent on a woven. Below that, the seam is the weak point of the garment and it will fail before the cloth does.

Both tests take under an hour and they happen before the marker is made. A factory that cannot show you a seam test result is selling you a hope.

The 2021 satin run that split at the hip

In 2021 we ran 40 satin slip dresses for a UK label. During the fit session, the side seam at the hip opened on nine of the forty pieces.

Three things had lined up badly. The seam allowance was 6mm, the overlock was running at 12 SPI, and the thread was tex 27 on a charmeuse that had been graded tight at the hip. Nothing was technically out of spec, because there was no seam spec to be out of.

We re-sewed the nine pieces and shipped four days late. Since then every satin style gets 1.5cm at the hip, a French seam on the side, tex 24 thread, and a pull test before bulk. On a later reorder of the same style we measured a 3mm opening at the same load, which is where we want it.

What to ask for on your next spec sheet

Ask for the SPI, the thread tex, the needle size, the stitch type and the seam allowance, per seam rather than per garment. Ask for the slippage test result with the load and the method named. If a factory quotes a single SPI for the whole garment, they have not thought about the fabric.

Then keep the answer. A spec sheet that names these five numbers is the difference between a seam that holds for a season and a seam that becomes a claim. Our cut and sew program sets out what we handle in-house, the case studies page shows completed runs, the FAQ page covers construction questions, and the production process page maps the three stages where a seam spec gets locked.

A private label clothing manufacturer usually prints the care instruction last, and that is the whole problem. The label is the only part of the garment that tells a customer how to keep it alive, and it often gets written before anyone has washed the fabric five times. When the instruction is wrong, the customer follows it, the garment shrinks or bleeds, and the return lands on your brand rather than ours.

So what has to be on the label, and who decides? The brand decides the instruction, the factory supplies the tested performance behind it, and the printer puts it on. If nobody has run the wash test, the instruction is a guess with your brand name attached.

What a care label must carry in the EU and the US

A care label does two jobs. It states the fibre composition and the country of origin, and it gives the washing, drying, bleaching, ironing and professional care instructions.

In the EU, Regulation 1007/2011 governs the fibre composition and requires the wording in the language of each member state where the garment is sold. Care symbols are voluntary in the EU, but if you use them they have to follow ISO 3758.

In the US, the FTC requires care instructions that will not harm the garment when they are followed, and the fibre content falls under the Textile Fiber Products Identification Act. Symbols are allowed under ASTM D5489, and most US brands print words alongside them.

Reading the five symbols that matter

A washtub with a number is the maximum wash temperature, and a hand in the tub means hand wash only. A triangle means bleach is allowed, and a triangle with a cross means it is not.

A square with a circle inside means tumble drying is permitted, and the dots inside set the heat. A plain iron carries one to three dots for 110, 150 and 200 degrees Celsius. A circle is professional care, with a letter inside naming the solvent family.

Most disputes we see are not about the symbols. They are about a brand printing a temperature the fabric cannot survive, usually because nobody tested it.

The five-wash test behind a care instruction

Care instructions should come out of a test, not a conversation. We run five home launderings to AATCC 135, or to ISO 6330 for a European market, then measure dimensional change and check appearance.

Five washes is stricter than the minimum. Three is the usual floor for a domestic claim, and we prefer five because a customer who likes a dress will wash it more than three times.

The numbers that come back set the instruction. Viscose moves 4 to 6 per cent, linen 2 to 3, cotton 2 to 4 if it is not pre-shrunk, and polyester under 1. If a fabric shrinks 5 per cent, you either pre-shrink it before cutting or you write a gentler instruction.

Why "dry clean only" is often a dodge

Dry clean only is legal, and sometimes it is the honest answer. A structured blazer with fused fronts, a coated fabric or a garment with a delicate trim all belong there.

It becomes a dodge when it is used to avoid a wash test on a fabric that would have passed. A customer who can put a dress through a 30 degree wash values it more than one that needs a trip to the dry cleaner, and price follows that.

We write dry clean only when the test supports it, and we say so when it is being used as a shortcut.

Where the label goes, and what it has to survive

The care label has to last the useful life of the garment, which is why a sewn-in woven label beats a printed heat transfer on a fine viscose. The transfer can lift after a few washes and take the instruction with it.

Placement convention is the side seam, five to eight centimetres above the hem, where a customer finds it without hunting. Some markets add requirements for the origin and the importer, so check the destination before the label is ordered.

Printed care labels usually carry a 500 piece minimum, the same as a printed brand label, so both artworks are normally approved in the same week.

The 2020 viscose run that shrank four centimetres

In 2020 we made a 60-piece viscose wrap dress for a US brand. The instruction on the label was taken from the fabric supplier's own note: machine wash warm, tumble dry low.

After the first home wash, 22 of the 60 came back with hems three to four centimetres short. The fabric had done what viscose does. The label had promised something the garment could not deliver.

We reprinted 60 labels at our own cost, offered a credit on the order, and changed the instruction to 30 degrees gentle with a flat dry. On the reorder we pre-shrank the fabric before cutting, which removed most of the movement.

The lesson is simple. A supplier's care note describes the fabric in the abstract. It does not know how your garment is constructed or what the customer will do to it.

How to lock the care code before bulk

Send the fabric with its shrinkage and colourfastness results before the label artwork is approved. Ask for the wash test report number and the date, because a report older than a year may not describe the current finishing line.

Then reduce it to three decisions: the maximum wash temperature, the drying method and the iron ceiling. Those three lines cover the instruction that actually gets followed.

Our private label program covers labelling and trims in one place, the case studies page shows completed runs, the FAQ page answers labelling questions, and the production process page shows where the care code gets signed off.

A high end clothing manufacturer keeps your approved sample after the order ships, and most brands never ask why. That one garment is the only thing that can settle a dispute about hem length, shoulder width or colour, because it is the version both sides agreed to. Without it, a bulk argument becomes two people comparing memories.

So what should a factory keep, and for how long? We hold seven items per style for two years, and we can tell you what goes wrong when one of them is missing. Ask any factory on your shortlist for its retention period and where the archive physically sits.

What we keep in the sample archive for each style

The centre of the archive is the signed approved sample, dated and photographed on the day it was approved. Everything else exists to explain it.

Next is the spec sheet with per-point tolerances: half a centimetre at the five key points, one centimetre elsewhere, and three millimetres on symmetry. Then the trim card, holding the actual zip, buttons, labels and thread tex that went into the approved piece.

Two fabric records follow. One is the fabric card with the dye lot number and a shade band cut from the roll. The other is the approved lab dips for every colour in the style.

The last two are the fit notes from the mock-up rounds and the marker record. Fit notes say what changed and why, which matters when a new pattern maker picks the style up a year later.

Storage is the unglamorous part of all this. Bins are labelled by style and season, the room is kept dry and out of direct sun, and the fabric cards live in a dark drawer because a swatch pinned to a wall fades inside a year. A faded reference card is worse than no card, because it argues for the wrong colour.

Pre-production, sealed and salesman samples are not the same thing

A pre-production sample is the last garment made before bulk, and it is the one that gets approved. The sealed sample is the signed copy of that pre-production piece, and it is what the archive holds.

A salesman sample is for selling and is not a production reference. Brands occasionally send one to a factory and ask for that fit, which is how a reorder starts drifting before it begins.

Ask which of the three you are looking at. It takes one question and prevents most of the confusion around approvals.

Why the approved sample beats a spec sheet in a dispute

A spec sheet gives you numbers. The approved sample gives you the version those numbers were measured on. When a hem is questioned, both sides measure the same garment instead of two different ones.

We photograph the approved sample with a tape measure at each of the five key points, and the PDF goes to the brand. A brand in New York can then check a shipment against the approved version without booking a flight.

Our Fashion Pains Solver, still in training, is being built for this exact step. The plan is that a brand uploads its five-point measurements and a photo, and gets a gap report against the approved sample before bulk ships.

The same file settles a colour question. The fabric card and the lab dip are the reference, and a photograph under a daylight lamp beats a description over email.

What it costs to skip all of this is easy to price. A third mock-up round runs $80 to $160 and five to ten working days. If the difference cannot be resolved by comparing samples, the remedy is a remake, and on 30 pieces at a five per cent rate that is roughly $40 in direct cost plus another week of schedule.

The 2019 run with no agreed hem tolerance

In 2019 we shipped a 30-piece crepe dress run with no written hem tolerance in the approval email. The brand measured the bulk hems 1.5 centimetres shorter than the sample they had approved.

We had measured to the same number, from a different starting point. Both readings were defensible, because nothing had been agreed about where the hem was measured from.

With no baseline, we settled it as goodwill: a credit and a re-hem at our own cost. Every approval email since then carries the five-point measurement table, and the brand signs the table rather than the sample alone.

What a reorder looks like when the archive exists

A reorder from a live archive starts from the approved garment and the paid-for pattern. There is no second pattern fee, no grading fee and no mock-up round, which is where most of the roughly 18 per cent saving on a second run comes from.

Without the archive, six months later the new sample drifts three to five millimetres and nobody can prove which version is right. That drift is usually invisible until the second run sits next to the first in a shop.

We ship a brand copy of the approved sample with the bulk and keep the archive copy here. Two copies, one set of numbers, and no argument about which one is the reference.

Four questions to ask any factory about samples

Ask whether the approved sample is kept, and for how long. Ask who signs off, and whether it happens in writing.

Ask whether the factory will photograph the five measurement points with a tape measure. Ask what happens when a bulk piece measures outside the tolerance, and who pays for the remake.

Ask for the retention period in writing on the order confirmation as well. A verbal promise to keep a sample is worth nothing once the person who made it has moved on.

A factory that cannot answer all four is not being careless. It simply has no archive, and every future order will be negotiated from scratch. Our high end production page lists the standards we hold, the case studies page shows completed runs, the FAQ page covers sampling questions, and the production process page shows where the sealed sample gets signed.

A high end clothing manufacturer should be able to name the womenswear categories it sews at 30 pieces, and the price of each. Most won't, because a short list is easier to defend than a long one. Here are ours: six categories, with the real per-piece cost at a 30-piece run.

What does 30 pieces actually cost per category? Between $45 and $140 FOB, and the spread is almost entirely construction rather than fabric. A slip dress and a lined midi dress use the same fabric budget and differ by 70 per cent on the invoice. Two of the six are the ones that surprise brands.

The prices below are FOB Guangzhou for 30 to 50 pieces, before freight and duty, in viscose, rayon, linen, cotton-linen, satin, polyester or TR suiting. If a category is not on this list, we say so on the first call instead of in week six.

Six womenswear categories, and what each costs at 30 pieces

• Slip dress or bias-cut slip: $52 to $70, 20 to 30 standard minutes, 1.6 to 1.9 metres of 145 cm fabric.
• Lined midi dress: $78 to $110, 40 to 60 minutes, 2.1 to 2.4 metres of shell plus 0.6 to 0.8 metres of lining.
• Occasion gown with a boned bodice: $95 to $140, 60 to 85 minutes, 2.6 to 3.2 metres of shell.
• Blazer in 240 to 280 gsm TR suiting: $78 to $105, 90 to 120 minutes, 1.8 to 2.1 metres plus 0.9 to 1.1 metres of lining.
• Jumpsuit: $68 to $95, 55 to 75 minutes, 1.9 to 2.3 metres.
• Skirt or simple two-piece set: $45 to $65, 25 to 45 minutes, 1.2 to 1.7 metres.

Those six cover most of a first womenswear collection. The two that get questioned on price are the lined midi and the blazer, and both questions are really about the same thing.

Why a lined midi dress costs twice a slip dress at the same quantity

Both are dresses at 30 pieces. The gap is 30 to 40 extra minutes and one extra fabric, and neither of those is the largest part of it.

Lining a dress means cutting a second set of panels, joining them, bagging the hem or attaching a facing, then pressing twice. On a 30-piece run that adds roughly 18 minutes per garment. At RMB 1.1 to 1.5 a minute, that is RMB 20 to 27, or $3 to $4.

The lining cloth adds $3.50 to $6.00 a garment. The rest of the gap is pattern work: a lined dress needs a lining pattern with its own ease, not a copy of the shell.

That last part is where we lost money in 2022. We cut a 30-piece lined midi with the lining pattern copied straight from the shell, one centimetre shorter at the hem. Twelve pieces pulled at the hem once the lining went in, and we re-cut the lining on all thirty.

It cost four days and about $210 in lining and minutes. Every lining block we cut now carries a two-centimetre hem drop of its own.

Fit checks that differ by womenswear category

Our base size is 8, which is EU 38 and IT 42, fitted on a model at 86-68-94 and 168 centimetres. Grading steps are plus 5 at the bust, plus 4.5 at the waist and plus 5 at the hip.

What gets checked at approval changes with the category. A slip dress is judged on the bust dart and the bias hang. A jumpsuit is judged on rise and torso length, measured from the shoulder-neck point to the crotch seam.

A blazer is judged on shoulder width within half a centimetre, armhole depth, and sleeve pitch. A skirt is judged on the hip and the hem line, checked on a mannequin from four sides.

Every approved piece is photographed with a tape at five points, so a brand in another time zone can check a shipment without booking a flight. A1, our design agent, is still in training on this step. The plan is that you photograph five points and get the deviation back before you sign.

What changes below 50 pieces for a womenswear brand

The fixed block is the same at 30 as at 300. Pattern, grading, two mock-up rounds, tech pack and marker come to roughly $1,100. Divided by 30 that is $36.67 a piece, and divided by 300 it is $3.67.

Three things get harder below 50. Colour count, because each extra colour needs a lab dip at $35 to $90 and its own marker. Fabric minimums, because dyed-to-order starts at 300 metres. Size splits, because five sizes across 30 pieces is six per size.

Our usual answer is two colours at 15 each, or three sizes at 10 each. Both keep the run at 30 without buying 300 metres of a single colour you will not sell.

How we quote a 30-piece womenswear order

Send a tech pack, or reference photos and your measurements, and a quote comes back within 24 hours. Fabric, cut-make-trim, trims, sampling and freight are separate lines, and the sampling fee comes off the bulk once you place the order.

Timing is two to three mock-ups in 14 to 21 days, a final sample in 28 to 42 days, and bulk in 6 to 10 weeks. If we source the fabric, add two to four weeks for dyeing.

Deposit is 30 per cent, balance 70 per cent against the inspection report and shipping documents. Three inspections happen per garment: inline, end of line, and pre-shipment.

What we need from you before the first pattern

Five things: your size chart or the size you want to fit, three reference images, the fabric you have in mind or a budget per metre, the retail price you plan to charge, and the date you want stock.

The retail price matters more than brands expect. A $250 dress and a $900 dress at the same 30 pieces need different fabric weights, different lining and different trim budgets, and the quote moves by about 40 per cent.

Our high end production page lists the standards behind each line, the case studies page shows completed womenswear runs, the FAQ page answers sizing questions, and the production process page shows the three stages in order.

A private label clothing manufacturer works from three documents: your tech pack, your pattern, and a bill of materials. The third one is the one brands forget, and it is the only page that names the exact fabric, thread and trim approved for your order. Here are the twelve lines we insist on.

What is a BOM, and why does it matter more than the tech pack? Because the tech pack describes the garment and the BOM names the parts. A factory can follow your tech pack perfectly and still ship a different fabric, a different thread and a different zip, and nothing in the tech pack will prove it.

What a bill of materials contains, line by line

Our template is twelve lines. Eleven of them describe a part. The twelfth decides whether the other eleven hold.

• Shell fabric: composition, weight in gsm, width, mill reference, colour code, dye lot.
• Lining: the same six fields, plus whether it is cut on the same grain as the shell.
• Fusible interfacing: weight, resin type, application temperature, and where it goes.
• Sewing thread: tex number, colour code, brand, and stitches per inch at each seam.
• Zips: type, length in centimetres, tape colour, pull finish.
• Buttons and hardware: size, material, finish, and the pull test figure in kilograms.
• Labels: main, care and size, with fold type and exact placement.
• Hangtags: dimensions, paper weight, string or pin, and where they attach.
• Polybags: thickness in microns, dimensions, and the suffocation warning if it applies.
• Cartons: dimensions, board grade, and pieces per carton.
• Packing method: fold dimensions, tissue, pin or clip, and whether the garment is bagged flat or on a hanger.
• Substitution rule: what may be replaced, by whom, and with whose written approval.

Line twelve is the one most BOMs leave out. It is also the line that gets read out in a dispute two months later.

How substitution happens without anyone deciding it

Nobody at a factory wakes up planning to swap your fabric. Substitution is usually a small decision made under time pressure by someone who is not in your email thread.

Three routes account for most of it. The named fabric is out of stock and a similar one is on the shelf. The named mill has a different lot that reads the same on paper. The named trim is discontinued and the nearest equivalent is bought without a note.

A BOM closes all three by naming the mill reference and the dye lot, not just the colour. "Navy satin" is a description. "Navy satin, 120 gsm, 150 cm, mill reference 4471, lot 2209" is an instruction.

We learned that the expensive way in 2021. A 30-piece satin dress BOM said "satin, navy" and nothing else. The mill was short, the factory bought from a second supplier, and nine of thirty pieces came out with a bluer cast, about 2.3 Delta E apart.

We settled it with a credit and re-cut the nine. The BOM template has carried a mill reference and a lot number on every fabric line since.

The three documents that are not a BOM

A tech pack is the drawing and the construction. A pattern is the shape. A trim card is a physical board of the actual parts, which comes closest to a BOM and is still not the same document.

We keep all four, and the BOM is the one that gets signed. The trim card shows you what a zip looks like. The BOM tells you which zip, from which supplier, in which length, on which garment.

Brands sometimes send a trim card and call it a BOM. It photographs well and it answers none of the six fabric fields.

Who owns the BOM, and what happens when you change mills

You own it. We write it, we send it with the quote, and you sign it before the fabric is ordered. If you change mills, the BOM changes with a revision number and a date, and the old revision is archived.

That sounds bureaucratic for a 30-piece order. It takes about twenty minutes, and it is the cheapest insurance in the whole process.

The BOM also survives a change of factory. A new supplier can quote straight from it, which turns a switch from weeks of back-and-forth into a few days.

What a swap actually costs on a 30-piece private label run

Take a 30-piece satin dress at $72 FOB. The shell is $4.20 a metre and the run needs about 70 metres.

A factory that buys a similar satin at $3.60 saves $42 on the order, which is $1.40 a piece. If the substitute is half a shade off, and the second roll is a different lot, you end up with two colours inside one run.

Twelve pieces that do not match is not a $42 problem. At a $26.50 remake cost per piece it is $318, plus a week of schedule, on a $2,160 order.

The same arithmetic applies to thread. A tex 27 thread swapped for a tex 24 to save $0.30 a cone shows up as a seam that fails the pull test at 6 kilograms, six weeks after the goods land.

How to sign off a BOM in ten minutes

Check the six fabric fields first: composition, weight, width, mill reference, colour and lot. If any of the six is blank, that line is a description rather than a specification.

Then check the thread tex and the zip length, because those two are where the word "equivalent" gets used most often. Then read line twelve out loud and ask what happens if the mill runs out.

If the answer is "we'll find something close", you have no BOM. You have a wish.

Our private label program sends the BOM with the quote as standard, the case studies page shows runs built from one, the FAQ page covers trim questions, and the production process page shows where the BOM gets signed against the approved sample.

A cut and sew manufacturer should inspect your fabric before the spreader moves, not after the first bundle is sewn. We score every roll on the four-point system, and roughly one roll in twelve fails. Here is what the score means and where the line sits.

What is the four-point system, and why does a 30-piece order need it? It is a scoring method that turns "this roll looks fine" into a number you can argue with. On a small run one bad roll is not an inconvenience. It is a third of your order.

What the four-point system actually scores

An inspector walks the roll and scores each defect by its length. Up to three inches is one point, three to six inches is two, six to nine is three, and anything over nine inches is four.

No single linear metre can score more than four points, which stops one catastrophic flaw from sinking a whole roll. The total is then converted to points per 100 square yards.

First quality usually means 20 points or fewer per 100 square yards. Above that the roll is second quality, and the price you paid was a first-quality price.

What gets scored: slubs and thick places, missing picks, holes, oil and rust spots, shade bars, and skew. The inspector marks each one on the face of the fabric with tape so the cutter can see it later.

Where the rejection line sits for a small fabric order

We inspect 100 per cent of rolls on orders under 300 metres, because on a 30-piece run you cannot afford to find a bad roll in week three. Above 300 metres we move to a 10 per cent sample plus every roll that looks wrong at the end.

Our line is 20 points per 100 square yards, plus three hard stops regardless of score. Any hole stops the roll. Any shade bar longer than 15 centimetres stops it. Any width variation beyond two per cent stops it.

On a 120-metre order of viscose, that line rejects roughly one roll in twelve. That is not a high failure rate for the industry. It is a normal one that small orders rarely measure.

Four checks that happen before the four-point score

Width first, measured at three points along the roll. A 145-centimetre roll that measures 142 changes your marker and your yield before a single panel is cut.

Then weight, by cutting a square metre and weighing it. A 120 gsm fabric that arrives at 108 behaves differently on the body, and a dress that looked opaque on the sample can go sheer in daylight.

Then shade, by cutting a band from the head, middle and tail of each roll and comparing them in daylight. Two rolls from the same lot can still differ by one to two Delta E, which is visible on a solid colour and invisible on a print.

Then skew, by pulling a thread across the width and measuring how far it runs off square. Anything above two per cent on a plain fabric twists after washing, and no amount of pressing fixes it.

What happens when a roll fails inspection

Three outcomes, and all three are written into the fabric contract before the mill ships.

The roll goes back and is replaced, which costs two to three weeks. The roll is accepted with a credit, usually 10 to 30 per cent of its value, and the usable metres are marked. Or the order is short-shipped and you cut fewer pieces.

What we do not do is cut it and hope. A defective roll that gets spread becomes defective garments, and by then the argument is about who owns thirty dresses instead of one roll.

Why fabric inspection changes your cutting plan

Every rejected roll changes the marker. If you planned 70 metres for 30 dresses and 6 metres are marked out, you are cutting 27.

That is why we inspect before the spreader starts rather than alongside it. The cut plan is built from inspected metres, not ordered metres, and the difference goes into the quote as a note rather than a surprise.

On a 30-piece run a 5 per cent allowance for defects and end-of-roll waste is normal. On 300 pieces it drops to 2 per cent, because the loss spreads across more garments.

We stopped trusting mill reports in 2020. A 120-metre viscose order arrived with a passing inspection report, so we skipped our own. Roll three had nine defects in its last fourteen metres, we were six metres short on the cut plan, and four dresses were never made.

We replaced them from stock fabric two shades off and sold those four at cost. Every roll has been inspected here since, including the ones with a certificate in the carton.

Three numbers to put in your fabric contract

Put the acceptance score in writing: 20 points per 100 square yards, first quality. Put the short-ship tolerance in writing, usually 3 to 5 per cent. Put the claim window in writing, usually seven days from delivery.

Without those three lines, the mill's own report is the only record, and it will say the roll passed.

Our cut and sew service includes incoming fabric inspection as a standard step, the small batch program explains why 100 per cent inspection is affordable below 300 metres, the case studies page shows runs that survived it, and the FAQ page covers claims.

Every small batch clothing manufacturer eventually has to explain printing, and the explanation starts with a number: 300 metres. That is the rotary screen minimum at most mills in Guangdong, and 500 at the larger ones. A 30-piece run needs about 68 metres, so screen printing is off the table before you pick a colour.

Digital printing has no such floor. Most mills will run 10 to 30 metres digitally, and a few will take five at a sampling rate. So the real question is not whether 30 pieces can be printed. It is whether the print cost per dress still works at your retail price.

Digital print minimums vs rotary screen minimums

Rotary screen is an industrial process. The mill engraves a screen for every colour, and that engraving costs $90 to $160 per screen before a metre is printed.

Then comes the run minimum: 300 metres at most mills, 500 at the bigger ones. At 500 metres a three-colour print lands around $3.60 to $4.80 a metre, which is why the price looks excellent in a quotation for two thousand dresses.

Digital reactive printing dyes the cloth rather than laying ink on top of it, so it behaves like the base fabric after washing. Minimums are five to thirty metres, and the price runs $6.80 to $9.20 a metre on viscose, linen or cotton.

Sublimation is the cheap digital route at $4.20 to $6.00 a metre, but it only bonds to polyester. On a viscose dress it will not fix, and the print washes out in patches within ten cycles.

What a 30-piece print run costs, line by line

Take 30 dresses at 2.1 metres each. That is 63 metres, plus 8 per cent for the marker and end-of-roll waste, so 68 metres.

A stock print, bought as already-printed cloth, costs $4.90 a metre with no print minimum: $333 for the run, about $11.10 a dress.

Digital reactive at $7.60 a metre is $517, plus $80 to $150 to separate the artwork. Call it $617, or $20.57 a dress.

Rotary screen looks cheapest per metre and is not. Three screens cost $360, the mill will not cut below 300 metres, and 300 metres at $3.90 is $1,170. That is $1,530 to print 68 metres and store 232.

Rotary only wins above roughly 420 pieces, where the stored metres become your reorder instead of dead stock. Below that, you are buying cloth to satisfy a machine.

One more number belongs in this table: the artwork. A repeat print needs a file that tiles cleanly, and a mill will charge $80 to $150 to fix one that does not. That fee is the same at 30 pieces as at 3,000, which is another fixed cost the unit price hides.

Colour count is the lever you actually control. Every colour removed from a screen print cuts a $90 to $160 engraving and a screen on the machine. Three colours is the practical ceiling for a small brand, and two is better.

Three print questions to ask before you send artwork

First, what is the base cloth? Sublimation needs polyester, reactive digital needs a cellulose fibre such as viscose, linen or cotton, and screen works on almost anything but needs volume.

Second, is the design a repeat or a placement? A repeat has to tile without a visible break, to within half a millimetre, or the seam shows on a 60-metre roll and the fault looks like our cutting.

Third, who approves the colour? A print strike-off is not a lab dip. Ask for it on your actual base cloth rather than on paper, and budget two rounds at $60 to $120 each.

Where print goes wrong on a small run

Registration drift comes first. Viscose stretches 2 to 4 per cent under tension, so a three-colour print can shift 3 millimetres across one roll, and the seam between panels shows it.

Hand-feel comes second. Pigment print sits on the surface of the cloth, so a pigment-printed slip dress feels like paper and cracks after twenty washes. Reactive print feels like the fabric because it is the fabric.

Crocking comes third. Rub the print with a white cloth for ten seconds. If the cloth greys, the print will mark a handbag in July.

We now quote a five-wash test on every print order, because a print that passes on day one and fails on wash three is not a print defect. It is a returned order with your label in it.

How to run 30 pieces of print without dead stock

Four routes work at this size. Buy a stock print and accept the mill's design; print digitally and pay the metre price; print cut panels instead of a roll; or put two styles on one print so the minimum is shared.

Panel printing is the route most brands never hear about. You send the marker, the mill prints the shapes, and waste drops to 12 to 18 per cent instead of a whole roll. It only works for a placement design, not an allover.

Our design agent A1 is still in training on print files, so a person checks every repeat tile before it reaches a mill. That check has caught three broken tiles this year, all of them invisible on a laptop screen.

In 2021 we quoted rotary screen on a 40-piece order and assumed the mill would waive its 300-metre minimum for an old customer. It did not. The brand paid for 300 metres, used 84, and 216 metres sat on our rack for eleven months until we sold it at cost.

Every print quotation from us now shows the minimum as its own line, with the stored metres calculated, so nobody discovers it after the deposit has cleared.

The small batch program prices print separately for exactly this reason, the case studies include a run where we changed print method after the first mock-up, and the FAQ covers strike-offs and print claims.

A high end clothing manufacturer can inspect its own work and still miss what a buyer sees at arm's length. That is why third-party inspection exists, and why we recommend it on a first order rather than every order. Nobody in the chain is neutral when the factory is paying for the report.

One man-day in Guangzhou costs $250 to $350. On a 30-piece run that is $8 to $11 a garment, which is either cheap insurance or 12 per cent of your FOB, depending on what the order is worth to you.

What a third-party inspection costs and covers

The fee buys a man-day, not a guarantee. Travel outside Panyu or Zhongshan adds $40 to $80, and booking between August and October needs seven to ten days' notice because every brand wants its check before the same shipping window.

The report runs four to six pages: quantity and carton check, a measurement chart against your spec points, defect classification, packing and label check, and photographs of every defect found.

The inspector opens a sample, not the lot. That is the part brands misread, and it is why a passed inspection is not a promise about every garment in the box.

AQL 2.5 on 30 pieces: eight garments, one defect

Under ISO 2859-1 at general level II, a lot of 26 to 50 pieces draws a sample of eight. At AQL 2.5 that sample accepts zero major defects and rejects the lot at one.

So a single puckered seam out of eight garments fails thirty. That sounds harsh until you remember the inspector cannot open the other twenty-two, and the factory knows which eight were opened.

Critical defects are always zero, at any AQL level: needle damage, a wrong care label, a measurement outside the tolerance table, anything that makes the garment unsellable.

Some brands move to AQL 4.0 so that one defect passes. On a lined garment we would not, because the defects that survive 4.0 are the ones a customer photographs and sends back.

When to pay for a third party, and when not to

Pay for it on a first order with any factory, on an order above $3,000, on anything with a stockist date you cannot move, and whenever part of the work is sub-contracted to another workshop.

Skip it on a reorder of a style that already passed, and on a 30-piece run where you approved the PP sample and receive inline photographs every day. At that size the fee can reach 10 per cent of the order.

There is a third case, which is a factory that refuses. That is not a reason to skip the inspection. It is the reason to book one.

What the report cannot tell you

It measures against the spec you supplied. If your spec says nothing about hem length, the inspector uses the factory's sheet, and the argument starts from their number rather than yours.

It says nothing about fit or drape. A garment can pass AQL 2.5 with a shoulder sitting 8 millimetres off, because nobody put the shoulder in the spec.

It samples. Eight garments out of thirty means twenty-two were never opened, and a good factory will tell you which eight it would have chosen.

What it also cannot tell you is whether the factory will fix what it finds. Ask that before you book, because a report with no remedy is a photograph of a problem you now own.

In 2021 a client booked an inspector without sending our tolerance table. The inspector measured hems against the factory's own sheet, failed six of eight pieces, and we spent four days proving they sat inside the spec we had signed.

The same year our own inline check passed nine satin dresses with puckered princess seams. The third party found them in twenty minutes. We reworked all nine in three days and paid for the reinspection ourselves.

How to brief an inspector so the report means something

Send five things with the booking: the spec sheet with every tolerance, photographs of the approved sample with a tape measure laid on it, the packing requirement, the label artwork, and the name of someone who answers within two hours.

Without the spec sheet, the report grades your order against someone else's standard. With it, the report is evidence you can act on, and the difference shows up the first time a lot fails.

Who pays, and what happens when the lot fails

The buyer pays the inspector, always. A factory that offers to pay is offering to choose the inspector.

Book the inspection for the day the goods are packed, not the day they are sewn. An inspector who arrives before packing can only check garments, while one who arrives after can check cartons, labels, polybags and the count as well.

When a lot fails there are three exits: rework and reinspect, ship with a credit, or cancel before the balance. Reinspection usually costs more than the defect, so we push for rework inside the same week.

Our own numbers are 3 to 8 per cent rework on a first run and 1 to 3 per cent on a reorder. An inspection that finds the 3 per cent early is cheaper than a claim that finds it at your stockist.

Our own rule is blunt: any first order above $2,000 gets checked by someone who does not work for us, whether the brand asks for it or not. We would rather be checked than trusted.

Our production standards are written to be inspected against, the case studies show what a failed lot looks like in practice, and the FAQ covers who pays for rework.

Most private label clothing manufacturer disputes we have seen were not about quality. They were about a document nobody wrote. When the spec is silent, the argument is settled by whoever is more stubborn in week nine.

A two-page production agreement, signed before the deposit, answers those questions in advance. Here are the nine clauses we will not start without, and the specific failure each one prevents.

The nine clauses a production agreement must contain

Parties first. The legal name must match the bank account exactly, with the registration number beside it. A payment sent to a name that does not match the invoice is how brands lose money to a lookalike account.

Scope second: style numbers, size ratio, and quantity with a tolerance. Price and incoterm third: FOB or DDP, currency, and what the price excludes, such as duty or testing.

Payment fourth: 30 per cent against a proforma invoice, 70 per cent against documents. Delivery fifth, and it needs a definition rather than a date on its own.

Quality sixth: the approved sample as reference, an AQL level, and a tolerance table attached as an annex. IP seventh: patterns, artwork and labels, and who holds each one after the run ends.

Claims eighth: a window, the evidence required, and the remedies. Cancellation and law ninth: what happens to the deposit at each stage, and which law governs if it goes wrong.

What quantity tolerance means on a 30-piece order

Fabric arrives short, rolls end early, and defects get cut around. A 3 per cent tolerance on 30 pieces is 0.9 of a garment, so the practical wording is 30 pieces, plus or minus one.

The invoice follows shipped quantity, not ordered quantity. If you need exactly 30 for a stockist, put that in the agreement and accept the cost of cutting a spare.

Over-shipment is not a gift. Either it is billed at the agreed price or it is a free sample, and the agreement should say which before the carton is packed.

Delivery date: ex-factory, on board, or delivered

Three different dates hide inside the word delivery. Ex-factory is when the goods leave our floor, on board is when the aircraft or vessel loads them, and delivered is when they clear customs at your address.

Put the leg in writing: five to seven days by air, 28 to 35 by sea, plus two to five days of customs at either end. A date without a leg is a wish with a deadline attached.

Force majeure should name what counts: a typhoon closure at the port, a scheduled power cut, Chinese New Year, a mill failure. Name the effect too, which is an extension of the date rather than a cancellation.

Add a notice clause while you are there: two working days to report a delay, in writing, with the reason. Most late orders are known to be late in week two and reported in week eight.

Cancellation, deposit and the clause everyone skips

Write the forfeiture schedule before you need it. Before fabric is ordered, the deposit is refunded minus pattern work, usually $180 to $260. After the cloth is bought and cut, the deposit is gone, because the cloth is already yours.

On a $2,400 order the 30 per cent deposit is $720, which is roughly what the fabric costs. That is not a coincidence, and it is why a factory cannot refund it once the spreader has run.

A verbal we will sort it out is what turns into a three-week argument with a lawyer's letter at the end of it.

Ask one question about the deposit: what exactly is it buying? If the answer is a percentage, ask again. On a 30-piece run the deposit buys fabric and trim, and a factory that cannot name the metres has not costed your order.

Why two pages beat a twelve-page template

A lawyer's template costs $300 to $600 and cites Incoterms revisions and arbitration rules a 30-piece order will never reach. What actually decides the outcome is names, numbers, dates, tolerances and ownership.

Nine clauses fit on two pages. Ours is written in plain English with the technical annexes attached, because a document nobody reads protects nobody.

Language matters more than length. If the agreement is in English and the workshop reads Chinese, attach a one-page Chinese summary of the nine items. Ours has one, and it is the page the cutting room actually uses.

In 2020 we shipped a 40-piece order with no written hem tolerance. The brand measured the hem 1.5 centimetres shorter than our sample, neither side could prove where the measurement started, and we settled with a $310 credit.

Two pages and a measurement chart would have cost nothing. Every agreement we sign now carries the annex, and so does every quotation we send.

What to attach to the agreement

Attach four documents: the spec sheet with tolerances, photographs of the approved sample with a tape measure on it, the trim card, and the fabric card with mill reference and lot numbers.

Those four annexes are what a claim is argued from later. Without them the agreement is a promise; with them it is a record.

Fashion Pains Solver, our claim-tracking tool, is still in training, so claims currently run through a spreadsheet and one named person. That is slower than it will be, and faster than an email thread nobody owns.

Our private label program sends a draft agreement with the first quotation, the case studies show how a tolerance annex settled a real dispute, and the FAQ answers the deposit questions this document exists to prevent.

Six millimetres is what separates an approved sample from a bulk garment, and a high end clothing manufacturer measures for it before the marker is drawn. Fabric arrives holding whatever moisture it picked up in transit. A roll cut the day it lands will not measure the same 48 hours later, so every roll here is conditioned before it goes near a table.

The question we hear most is simpler than that: why does bulk measure differently from the sample I signed off? Because the sample was usually cut from cloth that had already settled in a warm room, and bulk is often cut from cloth that came off a truck that morning.

Fabric regain: why a roll cut on arrival measures short

Regain is the moisture a fibre holds at a given humidity, counted as a percentage of its dry weight. Viscose sits near 13 per cent, linen 12, cotton 8.5, polyester under 1. A 30-metre roll of viscose therefore carries close to 4 kilograms of water that nobody invoiced you for.

Move that roll into a heated room at 30 per cent humidity and the water leaves. The yarn contracts, the panel gets shorter, and the loss shows up at the hem, which is the first place a buyer measures.

It runs the other way too. Cut in a humid Guangzhou August and ship to a dry Milan winter, and the garment arrives a size smaller than the one you approved.

The conditioning room: 20°C, 65 per cent humidity, 24 to 48 hours

We relax every roll in a room held at 20 degrees Celsius and 65 per cent relative humidity, give or take four. Woven cotton and polyester need 24 hours. Viscose, linen and their blends get 48, and we turn the rolls once.

That room holds about 400 metres and costs rent nobody can bill to a client. It also adds one to two days to every lead time, which is exactly why a factory chasing a fast quote skips it.

Ask any supplier how long they condition fabric and listen for a number of hours. A vague answer means the roll went from the delivery bay to the cutting table in the same afternoon.

Viscose, linen and polyester shrink on three different curves

Viscose is the one that surprises first-time brands. Without compaction at the mill it can lose 4 to 6 per cent of length on the first home wash, and no amount of resting fixes a fabric that was never pre-shrunk.

Linen moves 3 to 5 per cent, usually across the first two washes, then settles. Polyester moves under 1 per cent, and TR suiting sits near 1.5. We ask the mill for the compaction figure in writing before we quote a viscose style, and we decline the order if they cannot give one.

Container sweat: the fabric that arrives wet

Sea freight from Guangzhou to Rotterdam runs 28 to 35 days, and a steel container crossing climates condenses water on the inside of the roof. Rolls packed in poly bags arrive with damp patches on the outer layer and a smell that only shows up after pressing.

We open every carton on arrival and air the rolls for a day before inspection. Twice in the last three years we rejected 60-metre lengths at the outer edge and asked the mill to replace them, which they did.

The 50-centimetre wash test we run before quoting

We cut a 50 by 50 centimetre square from the roll, measure it, wash it at 40 degrees, dry it flat and measure again. The percentage goes into the quotation, and the square stays in a bag with the lot number on it.

That square has settled more arguments than any email. When a client says the bulk shrank, we can show what the cloth did before we ever cut a pattern.

Why the sample room and the cutting room must match

Our sample room and our cutting room share the same air handling, because a pattern cut at 22 degrees and sewn at 31 degrees gives two different garments. In July the difference between an air-conditioned office and an open workshop floor is nine degrees and 30 per cent humidity.

Small workshops rarely control this, and it is one reason two factories can sew the same pattern to visibly different lengths.

What goes into the tolerance table

Our table for a woven dress reads plus or minus 6 millimetres on body length, 5 on chest and waist, 4 on hem, and 5 on sleeve. Every number is measured on a garment that has rested flat for 24 hours after conditioning.

Without a table, a 6-millimetre difference becomes an argument about who is holding the tape. With one, it becomes a pass or a rework, decided in a minute.

Two runs where we cut too early

In 2022 we cut a 40-piece viscose order the same afternoon the fabric arrived, because a stockist date was tight. Nine dresses measured 8 millimetres short at the hem against the approved sample, and we recut them over three days at our cost.

In 2019 a linen blend went through the sample room in February with the heater running and the humidity at 32 per cent. The panels came out a centimetre narrow after the first wash, and the client found it before we did.

Both mistakes came from the same habit: trusting the calendar more than the cloth.

What this adds to your lead time and your price

One to two days, and roughly $0.20 to $0.40 per garment in handling that we absorb rather than bill. On a 30-piece run that is under $12 in total, against a hem rework that starts at $180.

Our production standards are built around conditioned cloth, the case studies show a hem dispute that a tolerance table settled, and the FAQ covers who pays when bulk measures off the sample.

A small batch clothing manufacturer spends nearly the same cutting hours on 30 pieces as on 300, and the cut ticket is the document that shows why. Thirty dresses in three sizes and two colours become four lays, and each lay is spread, marked, cut and bundled by hand.

So why does cutting cost so much on a short run? Because the labour is per lay, not per piece. Four lays at 30 pieces take close to the hours of ten lays at 300, while the fabric lost at the end of each marker is a far bigger share of a smaller order.

What a cut ticket actually contains

Eleven fields, and a workshop that cannot show you all of them is cutting from memory. Style number, colour, fabric lot, size ratio, plies per lay, marker length, metres issued, panels per garment, cutter name, date, and the bundle numbers that follow the pieces to the line.

The metres issued is the field brands should check first. If it does not match the consumption on your quotation, either the marker changed or the fabric is wider than planned, and both change your cost.

Ply limits: why satin takes 40 and chiffon takes 80

Ply is one layer of fabric in a lay. A straight blade cuts a stack, and the taller the stack the more the blade deflects at the bottom, so the limit is set by how much a fabric slips and bends.

Our working limits are 40 plies for satin and crepe, 30 to 40 for viscose, 60 for cotton poplin, and 60 to 80 for chiffon, which is thin enough that the blade error stays small. Push satin past 40 and the bottom panels come out 3 millimetres wider than the top ones.

Marker efficiency: 82 per cent at 30 pieces, 90 at 300

The marker is the nesting plan of every panel on the cloth. A big order can afford a marker that runs 12 metres and nests six sizes together, reaching 88 to 92 per cent efficiency. A 30-piece order nests three sizes and lands at 80 to 84.

That gap is real money. At 82 per cent a 1.55-metre dress needs 1.89 metres of cloth, and at 90 per cent it needs 1.72. On 30 pieces the difference is 5 metres, roughly $45 on a viscose at $9 a metre, and it is already inside the price you were quoted.

Two colours, three sizes: the four-lay arithmetic

Thirty pieces across two colours and three sizes is five per size per colour, which is too few to spread as one lay. We cut colour A in sizes S and M together, colour A in L on its own, then the same two lays for colour B.

Each lay is 5 to 10 plies, and the setup repeats four times: lay the paper, spread, check the grain, cut, notch, bundle. That repetition, not the sewing, is why a 30-piece run has a floor price.

What a lay costs in hours

A 12-metre marker spread 8 plies takes about 20 minutes if the cloth behaves and 40 if it does not. Cutting takes 15 minutes, notching and bundling another 25, and setup before the first ply is 30 to 40.

Four lays is therefore five to six hours of cutting room time for 30 garments. The same 30 garments inside a 300-piece order would take 40 minutes of that time, because one setup is amortised across ten times the pieces.

Cut edge and notching: the 3 millimetres that decide fit

The blade has thickness and it leans, so on a 40-ply lay the bottom panel can come out 2 to 3 millimetres wider than the top. That is enough to make a size 10 waist measure as a size 11 at the side seam.

We cut in two passes on satin and crepe and check the bottom ply against the marker before bundling. It is slow, and it is the difference between a graded run and a run where one size fits oddly.

Shade sorting: keeping one garment inside one roll

Rolls inside a single dye lot can still differ by half a shade at the edges, and a dress cut across two rolls can show a line at the waist seam. Where the marker allows it, we cut each garment from one roll and write the roll number on the bundle ticket.

On a 30-piece order we usually have three or four rolls, which is manageable. At 300 pieces the same discipline needs a shade band chart and a person whose only job is to enforce it.

The marker mistake that cost us three days

In 2021 we cut a marker planned for 44-inch fabric and the mill shipped 43-inch rolls, because the substitution was approved by phone and nobody updated the ticket. Twelve panels came out short at the selvedge and had to be recut from the balance.

Three days lost, and a delivery date we had to renegotiate with the client. The fix was a line on the ticket: fabric width verified at goods-in, signed by the cutter.

What to ask before your first cut

Ask how many lays your order will run in, what the ply limit is for your fabric, and whether you will receive the cut ticket with bundle numbers. Three answers tell you whether the factory planned your order or just accepted it.

Our small batch program sends the cut ticket with every first run, the case studies show a 40-piece order cut in five lays, and the FAQ answers the consumption questions this document creates.

Fashion Pains Solver, the tool we are building to log each cut and its bundle numbers, is still in training. Until it ships, the ticket is paper, signed, and photographed before the spreader moves.

A private label clothing manufacturer buys thread by colour rather than by style, and the seam is where a $600 dress starts to look like a $200 one. Thread is dyed to a shade card, never to your fabric, so a match is a decision somebody has to make on purpose.

The question that arrives most often is why the stitching reads lighter than the cloth. Two reasons: polyester thread reflects light differently from a matte viscose, and the match was approved under a shop light instead of daylight.

Tex numbers: the four weights a small brand actually needs

Tex is the weight of 1,000 metres of thread in grams, and it replaces the ticket numbers that mean different things in different countries. Tex 40 suits silk and lightweight viscose, Tex 60 covers cotton and linen shirting, Tex 80 handles suiting and denim, and Tex 105 is for visible topstitch.

Go one weight too heavy on a fine viscose and the seam puckers before the thread ever breaks. Go too light on a tailored jacket and the seam grins open at the armhole within a month of wear.

Dyed to match: what it costs and when it is worth it

A thread supplier's shade card is free and covers about 400 colours. Dyed to match your fabric is a different purchase: a minimum of 12 to 24 cones, 5,000 metres each, a lot fee of $80 to $160, and seven to ten days.

On a 30-piece run that is 60,000 metres of thread you will not finish, and it is still the right call on a $40 garment where the seam is meant to disappear. On a $12 cotton tee, a shade card match one step darker is the professional answer.

Why the same thread looks darker on satin and lighter on crepe

Light behaves differently on a flat filament surface than on a textured one. A navy polyester thread on navy satin picks up a specular highlight along the stitch line and reads lighter; the same thread on crepe sinks into the texture and reads darker.

Then the lighting changes the verdict. A 3000K shop light pushes everything warm, and a 5000K daylight lamp shows the difference you will see in a client's showroom. We approve thread under both, and photograph the seam.

Thread lot numbers: the banding nobody checks

Two cones of the same colour from different lots can differ enough to see when they sit in one garment. The dye bath changes between runs, and a supplier shipping from stock will not warn you.

In 2023 we finished 60 per cent of a 40-piece run with a second lot of navy that arrived mid-production. Eleven dresses showed a lighter seam under studio light, and we restitched them over four days. Every thread order now carries a lot number on the ticket, and we buy the whole run in one lot or not at all.

Contrast stitching: choose it on purpose

Visible topstitch is a design decision, not a matching failure. When a brand asks for contrast, we ask for a Pantone TCX number rather than a description, and we suggest two shades darker than the fabric for a seam that reads as intentional.

The stitch length matters as much as the colour. A 3.0 to 3.5 millimetre topstitch in Tex 80 reads as tailoring, while the same colour at 2.2 millimetres reads as a repair.

Tension and pucker: the line that looks like a stain

Too much needle tension pulls the fabric into the stitch and creates a puckered line that photographs as a dark stripe. It is not a colour problem at all, and brands regularly send us photographs of it described as a dye fault.

We sew a 30-centimetre test seam on the actual cloth before every run and check both sides. It takes four minutes and it is the only way to set tension on a fabric that has not been on that machine before.

Colourfastness: the wash test before bulk

Thread can bleed even when the fabric does not, particularly on saturated reds and deep indigos. We wash a sewn sample at 40 degrees with a white cotton cloth and check whether the white picks up colour.

If it does, the thread changes, not the washing instruction. A care label that says cold hand wash is a workaround, and a client who ignores it will be the one who finds out.

How many cones a 30-piece run actually needs

A dress uses 90 to 140 metres of thread, so 30 dresses need about 4,000 metres, which is less than a single 5,000-metre cone. Most of the thread cost on a short run is the minimum order, not the sewing.

We still buy two cones, one for the machine and one for the overlocker. A cone that runs out mid-lay is how one seam ends up a shade off the rest of the garment.

Storage: why thread snaps after six months

Thread takes up moisture and loses strength in a hot, dry store, and a cone left on a sunny shelf for a season will break at the needle under normal tension. We keep stock sealed at 20 degrees and use it within 12 months.

Old thread is the quiet cause of seam failures that get blamed on the fabric.

What to send with your artwork

Three things: a Pantone TCX reference for the thread, a fabric swatch from the actual bulk lot, and one sentence on whether you want the seam to disappear or to show. Those three answers remove most of the back-and-forth.

Our private label program keeps the approved thread cone with the sealed sample, the case studies show a lot-change restitch, and the FAQ covers trim minimums on short runs.

A small batch clothing manufacturer holds your fabric for four to six weeks before a blade touches it, and in Guangzhou that means holding it through 34 degrees and 85 per cent humidity. Storage is not a warehouse problem. It decides whether the cloth that reaches the cutting table is the cloth you approved.

The question we get most often is whether fabric can be damaged before cutting. It can, and the damage is rarely visible on the roll. Mildew starts on an inner fold, fading shows only on the outer layer, and a roll stored upright for a month goes out of round and skews the moment it is spread.

How to store fabric rolls without flat spots

Rolls lie horizontal here, on three-level racks, four to a bay, 12 bays. Nothing stands on end. A 1.5 metre roll left upright for three weeks goes oval by 8 to 12 millimetres at the core, and an oval roll feeds unevenly through the spreader.

Four hundred rolls is about 6,200 metres, which at 30 pieces a style is close to three seasons of work. That is why the store is a room with a door and a lock, not a corner behind the cutting table.

Weight is the quieter problem. A 30 kilogram roll resting on a pallet for two months takes a set along its bottom third, and that flat section will not lie flat on the table however long you relax it. We rotate rack positions every three weeks.

Nothing sits on the floor. Pallets hold a 12 centimetre clearance, which sounds excessive until the drains back up.

What humidity does to viscose, linen and polyester

Viscose takes up about 13 per cent of its weight in water at 65 per cent relative humidity, linen about 12, cotton 8.5, polyester under 1. That gap is why a store that feels fine to you is not fine for the cloth.

We hold 20 to 26 degrees and 55 to 65 per cent relative humidity, with two 50 litre dehumidifiers running from May to October. Above 70 per cent for ten days, mildew risk climbs fast, and viscose shows it first.

Below 45 per cent the opposite happens. Cloth dries out, turns brittle at the fold, and static makes a satin roll cling to itself during spreading.

Linen is the harder case. It drinks moisture and then dries stiff, so a linen roll that sits through a wet week needs a day back at 55 per cent before it will spread without cracking along the fold.

Why the roll label matters more than the shelf

Every roll carries four things: lot number, metres received, date received, and the style it was bought for. Without the lot number you cannot match the two rolls that must be cut together, and a shade difference between panels is the defect a customer sees first.

First in, first out is the rule, and it is harder than it sounds when a new delivery looks tidier than the old stack. We write the received date on the end of the roll in marker pen, not in a spreadsheet nobody opens in the store.

What a Guangzhou summer actually did to us

In June 2019 we stored 60 rolls of viscose on the floor because the racks were full. A typhoon pushed three centimetres of water across the store overnight, and 18 rolls were rejected at inspection the next morning.

The loss was $2,100 in cloth plus four days of schedule. What changed: pallets under everything, a $60 water alarm at the drain, and a rule that the racks are never allowed to be full enough to justify the floor.

We also learned that water damage does not look like damage. The outer layer dries, the core stays damp, and the mildew appears nine days later on a roll that seemed fine.

Storage time is not free, and it is not unlimited

We hold cloth for 30 days at no charge, then $0.18 per metre per month. That is not a penalty. It is the cost of the space, the dehumidifier and the person who checks the store twice a week.

Cloth held past three months is not ruined, but it needs re-inspection before cutting, and re-inspection costs a day. If your launch is eight months out, the honest advice is to buy later rather than store longer.

Rolls that arrive damp, and the check we run first

Cloth shipped by sea can arrive with condensation inside the bag, especially if it was packed warm. We open every roll within 24 hours of arrival and run a hand down the inner fold, because that is where damp sits longest.

A damp roll goes straight to the conditioning room at 55 per cent, not onto a rack. Drying it slowly takes two days; drying it in the sun takes an afternoon and ruins the hand of the cloth.

Two deliveries in the last year arrived with a wet outer layer from a truck that had no cover. Both were fine after 48 hours of slow drying, and both would have mildewed if we had stacked them.

What to ask before you ship fabric to a factory

Four questions to ask. Where does it sit, on what, and at what humidity — plus who checks it, and how often. Ask what happens if a roll is damaged in store, and what the storage charge is after the first month.

A factory that answers all four without leaving the room has a store. One that answers "in our warehouse, it is fine" does not.

If you are supplying your own cloth, our small batch program covers the handover, the case studies show a roll-level rejection, and the FAQ covers storage charges.

A high end clothing manufacturer spends more on needles in a week than on thread in a month, and almost nobody asks about it. Needle size and point decide whether a seam sits flat or leaves a line of holes down the front of a dress.

The question behind most needle conversations is why the first three centimetres of a seam pucker and the rest of it looks fine. It is rarely the machine. It is a needle meeting cloth it was not chosen for, at the point where that cloth has been folded longest.

Needle size by fabric weight: the four numbers we keep on the table

70/10 for 19 to 22 momme silk and fine viscose. 80/12 for standard viscose, lining and satin. 90/14 for cotton, linen and cotton-linen blends. 100/16 for TR suiting and heavier twill.

The first number is the metric size in hundredths of a millimetre, the second is the old Singer number. A 90/14 needle is 0.9 millimetres across the blade, and that is the hole it makes.

One size too heavy is not a small error. On a fine viscose it puckers the seam before the thread ever breaks, and on satin it leaves a permanent track along the stitch line.

Sharp, ballpoint or cutting point

Woven cloth takes a sharp needle with a round shank and a conical point, which pushes between yarns instead of cutting them. That is our default, because almost everything we sew is woven.

Ballpoint has a rounded tip that slips between loops rather than piercing them, and it belongs on jersey and rib. A sharp on a knit is how a small hole becomes a run after three washes.

Cutting point is for coated cloth and leather-look finishes, where a sharp tip would tear the surface rather than part it. We use it rarely, and never on satin.

Why the first three centimetres pucker

The start of a seam sits on a fold that has been pressed into the roll for weeks. The weave there is compacted, so the needle meets denser cloth than it will ten centimetres later, and the feed dog pulls against a section that will not stretch.

Two fixes work. Start a centimetre inside the seam allowance and backstitch over it, or run a strip of scrap cloth through first as a leader. Both take seconds and both remove the pucker.

If the pucker runs the whole seam instead, it is tension or needle size, not the fold.

Needle plate and feed dog: the other half of the pucker

A correct needle still puckers if the feed dog is set for heavy fabric. Fine cloth needs a fine-tooth dog and a needle plate with a small hole, otherwise the cloth is dragged down into the gap between them.

We keep a second plate and dog set for silk and fine viscose, and swapping them takes about ten minutes. Most factories do not bother, because most runs are not silk.

Eight hours, then change the needle

An industrial needle is good for about eight hours of running, or roughly 4,000 to 6,000 stitches through medium cloth. After that the point dulls, the blade picks up a burr, and the burr cuts filaments instead of parting them.

Needles cost $0.12 to $0.30 each, so a 30-piece run spends maybe $4 on them across three machines. A dull needle spends that much in rework on the first garment it marks.

We change at the start of every shift, and immediately after any needle break, because a broken needle can leave a fragment inside the seam where nobody will find it until a customer does.

Needle holes that never close

On 19 momme silk, a 90/14 leaves a 0.9 millimetre hole and the fibre does not recover. Under a 5000K lamp it reads as a dotted line beside the seam, and pressing does not remove it.

In 2021 we ran 40 silk blouses with the wrong needle in one machine. Nine front panels carried a visible line along the princess seam, we remade them, and the run shipped five days late.

The cost was about $380 in cloth and labour. The cause was a needle nobody had changed, because the machine was sewing "fine".

The four-minute test seam

Before every run we sew 30 centimetres on the actual bulk cloth, in both directions, and look at it under daylight. We check for holes holding light, for pucker at the start, and for the stitch sitting in the cloth rather than on top of it.

That is four minutes against a remake that takes five days. It is the cheapest test in the building.

Why the same needle behaves differently on two machines

Two identical machines set up by two operators will not sew the same seam. Needle height, presser foot pressure and the timing of the take-up lever all change how the stitch is drawn into the cloth.

That is why we test on the machine that will run the bulk, not on the sample machine. The sample machine is set for one garment a day; the line machine is set for forty.

What to put in your tech pack about stitching

Three lines are enough: needle size, stitch density per centimetre, and whether the seam should be visible. Add the thread Tex number if you have one, and we will match the rest.

If you do not know the numbers, send the cloth and a sample you like. We will set the needle from the swatch and photograph the test seam before the run starts.

Our production standards page lists the needle schedule, the case studies include a needle-mark remake, and the FAQ covers who supplies needles on a short run.

A private label clothing manufacturer approves a sample on a form and then ships it to a founder who tries it on a person. The gap between those two moments is where most fit complaints are born.

The question is always the same: the sample fit the mannequin, so why does it pull across the back on my customer? A dress form has no posture, no asymmetry and no opinion about how the wearer stands.

Six forms, and why we keep two body shapes

Our sample room holds six forms: S, M and L in a straight shape, and the same three in a curvier hip. The measurements are printed on the neck of each one, because a form that has been re-padded is a form you cannot trust.

Straight M is 90 cm bust, 72 waist, 96 hip. Curvy M is the same bust and waist with a 102 hip and a lower waist drop. Same label, two different customers.

We also keep one form at 68 cm waist for archive samples from 2018, and nothing is fitted on it any more.

What a dress form cannot tell you

A form cannot sit down, raise an arm, or lean. It has no shoulder slope difference between left and right, which most people carry by half a centimetre or more.

It also has no bust apex asymmetry and no posture. A garment that hangs clean on a rigid torso can pull at the blade on a body that stands with one shoulder forward.

That is why we treat a form as a first check, not an approval.

Form measurements are not your size chart

A form is a body. Your size chart is a garment. The difference between the two is ease, and ease is the number that decides whether a dress skims or grips.

We measure the form, then measure the sample on the form, and the gap is the ease we built. On a fitted viscose dress that is 4 to 6 cm at the bust and 2 to 3 at the waist. On a satin slip it is closer to 2.

If your chart lists garment measurements only, we will ask for body measurements before cutting the block.

Why a re-padded form lies

Forms get recovered every few years, and one padded by hand can gain a centimetre at the bust without anyone writing it down. We measure all six every quarter and log the result.

A form that is 1 cm off turns a correct pattern into a recurring complaint, and the pattern gets the blame.

The fit session record: 11 fields per garment

Every fitting produces one page. Style, size, form or model, date, the three key measurements before and after, where it pulls, where it gaps, the correction, who approved it, and a photo reference.

Eleven fields sounds like paperwork until the second sample arrives and nobody remembers whether the waist came in 1 cm or 1.5.

The record lives with the sealed sample, and the sealed sample is what bulk is measured against.

Sizes S to L is not a size range

Three forms cover three sizes, and three sizes is not a range a brand can sell into. A customer who is a 12 in one label is a 14 in another, and the only defence is a chart with numbers attached to it.

We ask for bust, waist and hip at four points, plus the height the garment is drafted for. A dress built for 168 cm sits differently on 160, and the hem is the first place it shows.

If the range has to reach a 16 or an 18, we grade from the base block and fit the largest size on a form that matches it. Grading without fitting the end size is how a range looks right in the middle and fails at both ends.

When we fit on a live model instead

For anything with a waist seam, a jumpsuit or a bias cut, we book a live model. Two models, 90 minutes, $60 to $90 per session depending on the category.

A model gives us movement: sitting, walking, reaching. That is where a back neckline gapes or a sleeve twists, and no form will show either.

On a 30-piece run the session costs about $2 a garment. On the runs where we skipped it, the correction cost more.

Photograph three angles at three metres

We shoot front, side and back at three metres with a 50 mm lens at waist height, under the same light each time. A phone held at 30 cm distorts proportion and makes every fit look wrong.

The photos go into the fit record with the measurements, so a founder in another timezone sees what we saw rather than a description of it.

The jumpsuit we approved on the wrong form

In 2020 we approved a 36-piece jumpsuit on the 68 cm form because it was closest to the client's spec sheet. Her fit model measured 72 cm at the waist.

Fourteen pieces had to be let out at the side seam, which cost six days and about $260 in rework. The pattern was right. The form was wrong.

Since then, no sample is approved on a single form. We fit the base size on two shapes and write down the difference.

What to send us before the first fitting

Three things: your size chart with body measurements rather than garment measurements, one sentence on who the customer is, and any garment that already fits her well.

A well-fitting garment from another brand tells us more than a page of numbers, because we can measure the ease instead of guessing it.

Our private label program page covers the fitting stages, the case studies show a size-chart correction, and the FAQ covers who owns the fit block.

A cut and sew manufacturer can run a perfectly straight seam and still hand you a dress that ripples at the hip after one wash. The culprit is almost never the sewing. It is the overlock settings nobody wrote down.

Why does the side seam wave? Because the overlocker stretched the edge while it was trimming it. On a 110 gsm viscose, a differential feed set for jersey pulls the seam allowance 3 to 4 per cent longer than the panel beneath it. The seam then has nowhere to go, so it ripples, and one wash at 40°C sets the wave for good.

We see it on light wovens — viscose, cupro, fine cotton lawn — and almost never on the TR suiting we run for tailoring clients. The fabric tells you where to look before the machine does.

What overlock tension does to a 110 gsm viscose seam

A four-thread overlock carries two needle threads and two loopers. The needles make the line you see on the face. The loopers wrap the raw edge.

Needle tension too high puckers the seam along its length before anyone presses it. Looper tension too high curls the edge into a tube, and the allowance will not lie flat under the next operation.

Our starting values on 110 gsm viscose: needle 3.5 on a 0 to 9 dial, upper looper 4, lower looper 3, stitch length 2.5 mm at 12 to 13 stitches per inch.

Those numbers move for a 200 gsm ponte. They do not move because a client prefers a softer look, which is a conversation we have had more than once.

Differential feed is the setting that gets left behind

The differential dial sets how far the front feed dog travels against the rear one. Above 1.0 it gathers. Below 1.0 it stretches.

Jersey runs at 1.3 to 1.5 so the edge does not flare. Woven viscose runs at 0.8 to 0.9, or 1.0 when the cloth has body.

Failures happen at the changeover. A line finishes a knit job on Friday and starts a viscose slip on Monday with the dial still at 1.3.

At 1.3 on a 60 cm side seam, the overlocked edge comes out about 2 cm longer than the panel. That is the wave, and it is there before the first wash.

Four-thread overlock or five-thread safety stitch

A four-thread overlock joins and finishes in one pass. A five-thread safety stitch adds a chainstitch beside the overlock, which is stronger and roughly 30 per cent slower.

Lined bodices get four-thread at 5 mm width. An unlined underarm or a crotch seam gets the safety stitch.

Width matters as much as thread. Under 4 mm on a fraying viscose, the edge sheds in the wash. Over 6 mm on a 1 cm allowance, the overlock eats the seam line and the garment loses a size.

Pucker, wave or flare: three faults that look alike

A pucker runs along the stitch line and shows before washing. Tension is the cause, usually the needle.

A wave sits in the seam allowance and shows after washing. Feed is the cause, and the fix is a dial, not a new operator.

A flare opens the seam edge outward on a knit. That is too little differential feed, and it is the one fault that gets worse when you try to pull the fabric straighter.

Three faults, three dials. Diagnosing the wrong one costs a week and a remake.

Spun polyester or woolly nylon

Spun polyester is our default on wovens. It has little stretch and presses flat, which is what a woven seam wants.

Woolly nylon bulks up and covers a raw edge better, but it will not press flat and it reads as a ridge under a light satin.

We keep woolly nylon for knits and for unlined pieces where the inside is visible. It costs about 2.5 times spun polyester per cone, which on 30 pieces is still under $6.

Where the wave shows first

The hip is where a side seam carries the most curve, so a stretched edge has the least room to recover. That is why the fault is always reported at the hip and not at the waist.

Underarm seams wave when the overlock is run downhill into the curve at full speed. We slow to about 60 per cent through the armhole and the fault mostly disappears.

A wavy hem is usually a different problem. The edge was overlocked before the hem was pressed, so it was stretched and then set by steam.

What a seam allowance actually needs

A 1 cm allowance is our default for woven womenswear. Below 8 mm the overlock has nothing to wrap. Above 1.5 cm it adds bulk at the hip exactly where a fitted dress shows it.

On a lined garment we trim to 8 mm after overlocking so the lining does not print a ridge through the shell.

If your tech pack specifies 1.5 cm everywhere, expect a heavier seam at the waist than the sample showed. That is a spec decision, not a factory fault.

How to test an overlocked seam before bulk

Ask for a 30 cm strip of your bulk fabric, overlocked on both long edges, washed at 40°C and dried flat.

Measure before and after. More than 2 per cent growth means the feed is wrong for that cloth.

Then press it. A seam that will not press flat will not sit flat on a body, and final ironing does not fix a stretched edge.

The test costs one metre of fabric and 20 minutes. We now run it on every new fabric before the cutting ticket is issued, and the result goes on the operation sheet.

What an overlocked edge costs in minutes

One side seam takes about 12 seconds on a four-thread machine, plus handling. A 30-piece run with four seams each is roughly 25 minutes of machine time.

That is not where the money goes. It goes into the 20 to 30 minutes of re-setting every time the fabric changes.

A run with three fabrics carries an hour and a half of setup that no quote line shows, and it is the first thing a factory skips when it is busy.

The jersey setting we left on a viscose run

In 2021 we cut 44 viscose slip dresses in one colourway. The line had come off a 240 gsm jersey job and the differential feed was still at 1.4.

Eighteen pieces showed a 4 mm wave along the side seam after the client's wash test. We remade them from the balance of the roll, which cost five days and about $410 in fabric and labour.

The overlocker was not broken. The setting was inherited, and nobody had written down what it should have been.

Since then the differential value sits on the operation sheet beside the stitch length, and the line leader signs it before the first garment is sewn.

Why this shows up at 30 pieces and not at 3,000

On a large run the first 50 pieces are inspected and the setting is corrected before the rest is cut. At 30 pieces there is no rest.

That is the real cost of small quantity. Fewer pieces to absorb a setup error, so the setup has to be right the first time.

Our cut and sew program writes overlock settings onto the operation sheet before the first garment, the case studies include a seam correction, and the FAQ covers who signs off on a setting change.

A small batch clothing manufacturer cuts 30 pieces in two or three lays, and every lay is a chance for the knife to drift. The panels that come out wrong are almost always the bottom ones.

Why are my panels inconsistent? Because a straight knife bends. The blade is 8 inches long and unsupported below the presser foot, so it leans away from the line as it travels down through the stack. The top panel is cut by the blade at the top; the bottom panel is cut by the tip, which is somewhere else.

On a 12 cm lay, the bottom panel can finish 2 to 3 mm wide on a curve. At 16 cm it reaches 5 mm, which is half a size at the hem.

Why a dull blade cuts a different size at the bottom of the lay

A sharp blade slices. A dull blade pushes, and the operator pushes harder to compensate, which adds lean on top of lean.

We sharpen every 2 to 3 lays, or every 40 to 60 metres of cutting, whichever comes first. A blade costs $9 to $14 and takes four minutes on the bench grinder.

Each cutting room keeps six blades in rotation so a change never waits for the grinder. A 30-piece run with three lays uses three blades, about $35 of consumable.

The tell is the edge itself. A dull cut looks slightly polished and the layers separate at the corner rather than holding square.

Blade speed and fused edges on polyester

Our straight knife runs at 3,400 rpm. On a 100 per cent polyester satin, that speed generates enough friction to melt the edge as it cuts.

A fused edge will not fray, which sounds like a benefit until you feel it. It reads as a crisp ribbon, shines under light, and will not drape with the rest of the panel.

We drop to 2,600 to 2,800 rpm on polyester and cut through a sheet of tissue paper above and below the lay. The paper takes the heat instead of the cloth.

On viscose and cupro we keep the speed up, because those fibres do not glaze and a slow blade drags the weave instead.

Lay height: what 30 pieces actually stacks to

Thirty pieces across three sizes is 10 layers per size. That is either three lays of 10, or one lay of 30 if someone is in a hurry.

Ten layers of 110 gsm viscose is about 5 cm of stack. Thirty layers is 14 cm, and the deflection at the bottom of a 14 cm lay is 4 to 5 mm on a curved seam.

Our limits are 6 cm for satin, 8 cm for viscose and cupro, and 10 cm for cotton poplin. Above that, the lay gets split whether the schedule likes it or not.

Splitting costs two extra setups of about 40 minutes each. It buys a millimetre, and a millimetre at the hem is the difference between two sizes.

The six measurements we take on every lay

After cutting, we measure the top panel and the bottom panel at three points: waist, hip and hem. Six numbers per lay, tolerance 2 mm.

If the bottom panel is outside tolerance, the lay is re-cut before it goes to the bundling table. Once panels are numbered and bundled, the fault is already in the run.

The six numbers go on the cut ticket with the fabric roll number, so a defect can be traced to a lay rather than to "the cutting room".

Pins, weights and tissue paper

Pins distort a lay. Every pin pushes the layers apart by the thickness of the shaft, which on 10 layers is measurable at the seam.

We weight the marker instead, and use pins only at the corner of a curved piece where a weight will not sit. On satin we pin inside the seam allowance and nowhere else, because a pin hole in the body of the panel is permanent.

A sheet of tissue paper between the bottom layer and the table stops the blade dragging the lowest panel out of line, which is the quiet cause of a short hem that no one can explain.

When we re-cut instead of trimming

Trimming a panel that is 4 mm narrow at the hem changes the hem curve and the grain line at the same time. It looks fixed and it is not.

Our rule: re-cut when the error is over 3 mm on a curved edge. Trim when the edge is straight and the allowance can absorb it.

Re-cutting needs fabric from the same dye lot, which is why we hold 8 to 10 per cent extra on small runs. On a 30-piece order that is 3 pieces of fabric sitting in the store.

The satin skirt we cut with a tired blade

In 2019 we cut a 30-piece satin midi skirt in three lays. The third lay went through a blade that was two lays past its change, because the grinder was in use.

Twelve hem panels finished 4 mm narrow. We re-cut them from 14 metres of the same lot, which cost three days and about $270 in fabric and labour.

Nothing was wrong with the pattern or the marker. One consumable item was treated as optional.

Now the blade count is written on the cut ticket and a lay cannot start past the count without a signature.

What to ask about a factory's cutting room

Four questions cover it. How many lays for 30 pieces, when is the blade changed, does anyone measure the bottom layer, and is the table vacuum or static?

A factory that answers "one lay" is optimising its own time, not your size consistency. A factory that cannot answer the blade question does not track consumables at all.

Ask for the six measurements on one lay of your own order. It takes five minutes and it tells you whether the tolerance exists outside a document.

Our small batch production page sets out lay limits by fabric, the case studies show a re-cut, and the FAQ covers who pays when panels are short.

A private label clothing manufacturer holds three copies of your pattern: the paper master, the CAD file and the cutting copy on the table. A reorder only fits the same if all three carry the same version.

Why does my third run fit differently from my first? Because a correction agreed after the first fitting was applied to one copy and not the others, and bulk was cut from the copy that missed it. Nobody made a mistake at the machine. The filing failed.

This is the least dramatic failure in garment production and the most expensive, because it repeats on every reorder until someone opens the pattern bag and reads a date.

Where a pattern version gets lost: paper, CAD and the cutting copy

The paper master lives in the sample room because that is what the pattern maker holds while fitting. The CAD file lives on one machine, often one operator's desktop. The cutting copy sits in a sleeve at the table.

Corrections almost always land on the paper first, because the fitting happens in front of it. The CAD is updated later, or not at all. The cutting copy is printed from the CAD.

Three copies, three chances, and the gap between them is measured in centimetres at the waist.

What a 1 cm waist correction touches

A waist reduction of 1 cm is not one line. On a fitted bodice it changes the front, the back, the waist facing, the lining front and the lining back. Five pieces, and two of them are cut from a different fabric.

A 0.5 cm armhole change touches the armhole facing, the sleeve head, the sleeve, the lining and the back neck facing. Five or six pieces depending on how the block is split.

If the correction is applied to the shell and not the lining, the lining will be 1 cm larger than the shell. The garment still closes. It hangs wrong, and the customer reads that as cheap fabric.

Grading from the wrong base size

Our womenswear grade rules run 5 cm at the bust, 4 at the waist and 5 at the hip between sizes, from a base of M.

If the base block is version 1 and the size run is graded from version 2, every size is correct and the base is not. Size 14 lands about 1 cm off at the waist, which is exactly the size that gets returned.

Grading is arithmetic. It cannot detect that its input was the wrong file.

The five fields on our pattern change log

Every change gets one line: style code, version number, date, what changed and on which piece, and who approved it.

The line is written on the pattern bag in marker and in the header of the CAD file. A version without a log is a rumour, and rumours do not survive a staff change.

We keep the log with the sealed sample, so the approved garment and the file that made it are in the same place.

Never overwrite a file

Versions get a suffix: _v1, _v2, _v3. The cutting copy is exported read-only, and the pattern maker's working file is never the file the cutter receives.

Overwriting is the single most common cause of a reorder that fits differently, because the old version disappears without a trace and the new one looks authoritative.

A file with no version in the name has no history, and a pattern without history cannot be audited.

Pattern custody: who holds the master, in what format

Ask for the master in DXF, AAMA or ASTM format, or the native file from Gerber or Lectra. A PDF is a picture of a pattern. You cannot grade from it.

The paper master is 120 gsm card in a hanging bag with the style code on the outside. Ours are stored flat in a humidity-controlled room, because card that absorbs moisture grows and the pattern grows with it.

Ask what happens if you leave. Who keeps the block, what is handed over, and whether the graded set goes with it.

Why the third run is the one that breaks

Run one is made from the pattern that was corrected during sampling, so it is right by construction. Run two follows within a season and the same people remember it.