Every brand that has ever placed a first hoodie order has been warned about these mistakes.

Almost nobody explains what they actually cost.

That gap is the problem. “Don’t skip your tech pack” is advice you can nod at and ignore, because it sounds like housekeeping.

“A missing points-of-measure table means your QC team has nothing to inspect against, so they pass 100% of a run that is 2cm off across every size” is advice you act on, because you can see the money leaving.

This page closes that gap for nine mistakes.

Each one gets the same five-part breakdown:

What it looks like from your side, why it happens from ours, what it costs in cash, what it looks like on a cutting table in Sialkot, and the exact fix.

Then all nine collapse into one thing: the Pre-Order Gate, and an 18-point checklist you run before you release a purchase order.

Nothing goes to bulk until every box is ticked. That is the whole system.

One caveat worth stating up front. The cost figures below are modelled on 2024-2026 and are directional, not quotes.

Fabric prices, freight, and dye costs move.

The shape of the numbers, which mistakes are recoverable and which ones are not is the part that stays true.

The money view

What does each mistake cost?

The column that matters is the last one.

Some of these mistakes cost you a delay.

Some of them cost you the run.

Knowing which is which tells you where to spend your attention before you place the order.

Mistake Cost at 100 units Cost at 500 units Recoverable?
Weak tech pack $900-1,400 $2,600-7,500 Sometimes. Extra sample rounds if caught early. Scrap if caught late.
Chasing the lowest price $600-1,100 $3,000-5,500 No. Shows up as returns and reviews months later.
Ignoring fabric minimums $1,200-3,000 $1,200-3,000 Partly. You own surplus fabric usable if you reorder.
Rushing sampling $1,500-4,000 $7,000-20,000 Rarely. This is the one that takes whole runs.
No golden sample $0-3,500 $0-17,000 No. With no sealed reference, you lose the dispute by default.
Loose color/wash spec $400-1,800 $2,000-9,000 Partly. Off-shade units sell at a discount, not at zero.
Wrong sizing curve $700-1,600 $3,500-8,000 Partly. Dead stock is cash frozen, not cash burned.
No inspection $300-1,200 $1,500-6,000 No. Once the balance is paid and the carton has shipped, leverage is gone.
Trend-heavy, no core $2,000+ per drop $8,000+ per drop No. The cost is the reorder revenue that never arrives.

How to read this:

Figures model direct loss rework, scrap, discount, freight on a mid-weight 400 gsm french terry hoodie.

They exclude the cost of a delayed launch, which for a drop-driven brand is often larger than the manufacturing loss itself.

Mistake 01 · The root cause

Skipping or under-building the tech pack:

This is the mistake that causes most of the other eight. Fix nothing else on this page and fix this, and you will still cut your risk in half.

What it looks like:

Your sample arrives. The body is right.

But the hood is shallow, the pocket bag is a different shape from the one in your head, the drawcord tips are silver metal instead of matte black, and the cuff is snugger than you wanted.

Nothing is wrong, exactly. It just is not your hoodie.

You send feedback.

Round two arrives three weeks later with two of the four things fixed and one new thing broken.

You are now six weeks and $450 into a garment you have not approved.

Why it happens:

Because the factory received a mood board, a front-and-back sketch, and a sentence that said “heavyweight, oversized, boxy fit.”

Every one of those words is an opinion, not a measurement.

So the pattern master does the only thing they can do:

They fill the gaps with the house block.

That is the factory’s own base pattern, refined over hundreds of orders, and it is genuinely good for the factory’s idea of an oversized hoodie. Not yours.

Every unspecified detail defaults to a house standard you have never seen.

What it costs:

Caught at sampling, it costs sample rounds roughly $120-180 each plus 2-3 weeks of calendar, and calendar is usually the expensive part.

Caught after bulk cut, it costs the fabric.

Once a marker is laid and 500 fronts are cut, the pieces cannot be uncut.

At a $11.40 unit cost, a 500-piece re-cut is $5,700 of fabric and labor that no longer exists.

From the floor:

Here is what most brands do not picture: a tech pack is not read by one person.

It is read by four, in four different rooms, and each one takes a different page.

The cutting room reads the fabric spec and the consumption.

The sewing line reads the construction callouts: flatlock or overlock, single or double needle, chainstitch or coverstitch at the hem.

The trims store reads the BOM. And QC reads the points-of-measure table.

That last one is the trap. QC does not judge whether the garment is good.

QC checks the finished garment against the POM table and flags anything outside tolerance.

If your tech pack has no POM table, QC has nothing to check against, so they check the garment against nothing, and everything passes.

We have seen a full run pass internal QC while being consistently 2 cm off across every size, because the only measurements in the file were on the sketch, and nobody grades a sketch.

The fix:

Build the tech pack before you request a single price.

At minimum, it needs a graded POM table with tolerances (±1 cm on body measurements is standard), a bill of materials naming every trim by supplier reference, construction callouts per seam, color references by Pantone code, and label and packaging placement.

If you do not know how to build one, that is fine, but know that it is the deliverable, not the sketch.

Mistake 02 · The reframe

Chasing the lowest price per unit:

What it looks like:

You send the same brief to six factories.

The quotes come back between $9.20 and $16.80 for what appears to be the same garment.

You take the $9.20, feel clever, and save $760 on a 100-piece run.

Eight months later, your return rate is climbing, and a review says the hoodie “shrank weirdly and the inside went bobbly after three washes.”

Why it happens:

Because price is not an input. Price is an output of a spec.

When two quotes for “the same” hoodie differ by 45%, they are not the same hoodie; you just did not specify tightly enough to see the difference.

The gap almost always lives in the same handful of places, and none of them are visible in a photo:

  • Yarn type. Open-end yarn is cheaper than ring-spun and pills faster. Nothing on the hangtag tells your customer which one they bought, but their washing machine will.
  • Actual GSM versus declared GSM. A fabric sold as 400gsm can land at 340gsm. That is a 15% fabric saving for the factory and a completely different garment for you.
  • Pre-shrink treatment. Compacting costs money. Skip it, and your customer’s size L becomes a size M.
  • Seam construction. A single-needle hem instead of a coverstitch. A serged shoulder instead of a taped one. Each saves seconds per garment; seconds are the entire labor line.
  • Thread and trims. Polyester thread substituted where cotton was implied. Unbranded zips. Draw cords that fray at the tip because there is no metal or molded stop.
What it costs:

The cheap quote wins on the invoice and loses everywhere else: returns, replacement shipping, refunds, and a review profile that suppresses conversion on every future drop.

A $760 saving that produces a 12% return rate on a $68 hoodie has already cost you more than it saved, and the reviews stay up forever.

From the floor:

Ask any factory to break a quote into its parts, and you will get roughly this: fabric, trims, cutting, stitching, printing or embroidery, wash and finishing, packing, overhead, and margin.

Fabric is typically the largest single line on a hoodie, often around half of the ex-works cost.

So when a competing quote comes in 30% cheaper, arithmetic tells you where it came from.

Nobody finds 30% in the packing line.

The honest version of a cheap quote sounds like, “We can hit that number on 320 gsm open-end, no compacting, single-needle hem.”

That is a real garment, at a real price, for a real market.

The dangerous version is the one that quotes your spec at that number and quietly builds the other garment.

The fix:

Quote the spec, not the garment.

Send the identical tech pack to three factories and demand a line-itemized quote for fabric, trims, CMT, finishing, and packing not one number.

Then compare the lines. If one factory’s fabric line is 30% below the others, ask exactly one question:

What yarn, what GSM, and is it compacted or not?

The answer will tell you whether you found a better factory or a different hoodie.

Mistake 03 · The design constraint nobody mentions

Ignoring fabric minimums:

What it looks like:

You design a hoodie in a custom heather with a brushed back, in three colorways, at 50 units per color.

You approve the design, build the tech pack, and get quotes.

Then the factory asks which mill is holding your fabric, and the whole thing stops.

Why it happens:

Because you have two separate minimums and you only knew about one of them.

The garment MOQ is the factory’s number of finished pieces they will cut and sew.

Ours is 50 pieces per style for most apparel and 100 for headwear.

That number is friendly, and it is the one every manufacturer advertises.

The fabric MOQ is the mill’s, and nobody advertises it.

A mill will not put a custom color on a dyeing machine for a handful of kilos; the setup, the dye bath, and the machine time cost the same whether you run 40 kg or 400 kg.

So custom-dyed fabric carries a per-color, per-quality minimum, and it is usually measured in the hundreds of kilograms.

Do the arithmetic. A 400 gsm French terry hoodie consumes roughly 0.9-1.1 kg of fabric once you include cutting waste.

If your mill’s custom-dye minimum is 300 kg per color, that minimum is a garment order of roughly 300 hoodies in that one color. You wanted 50.

What it costs:

You get three doors, and you have to walk through one:

  1. Buy the minimum. You now own 250 hoodies’ worth of fabric you did not budget for. It is not wasted; it is prepaid if you reorder. If you do not reorder, it is a write-off sitting in a warehouse.
  2. Switch to a stock-service fabric. Mills hold standard colors in standard qualities, available in small cuts. You lose the custom color; you keep the timeline and the cash.
  3. Pay a short-run surcharge. Some mills will run below minimum for a premium. Expect the fabric price per kilo to jump meaningfully often enough to make the whole style uneconomic.
From the floor:

The reason this catches people is sequencing.

Brands design first and source second, because designing is the fun part.

In Sialkot, the sourcing conversation happens first, always, and it is short:

What is on the shelf, what has to be dyed, and how much of each you have to take.

The workaround experienced brands use is to design into the stock-service range for a first run.

Pick a fabric the mill already holds: a good ring-spun terry in black, off-white, and a mid-grey heather, and put your identity into the fit, the trims, the label, and the print.

Then, once you know the style sells, you dye a custom color on the reorder when the minimum is not a stretch anymore.

Your first drop does not need a bespoke Pantone. It needs to sell out.

The fix:

Run a sourcing brief before you finalize the design, not after.

Ask the mill or your factory three questions:

What is available from stock in this quality, what is the minimum per color for a custom dye, and what is the lead time on each?

Then design inside the answers.

Mistake 04 · The expensive one

Rushing the sampling ladder:

If mistake one is the root cause, this is the one that takes the whole run.

Sampling is where you buy certainty.

Skip a rung, and you carry the uncertainty into bulk, where it multiplies by your order quantity.

What it looks like:

You have a launch date. Sampling is eating the calendar.

So you approve a fit sample from photographs, skip the pre-production sample, and tell the factory to go to bulk because “the fit sample was basically perfect.”

Then 500 hoodies arrive in a fabric that behaves differently from the one the sample was made from, with a print that sits 2 cm lower because bulk printing uses a different platen and a hem that puckers because the bulk line ran a different machine than the sample room.

Why it happens:

Launch-date pressure, every single time. Sampling feels like a delay.

It is not; it is the part of the project where mistakes are still cheap.

The ladder and what each rung buys you

01. Proto sample:

First physical translation of the tech pack, often in whatever similar fabric is to hand.

Buys you: Is this garment even the shape I drew? Do not judge fabric or color here.

02. Fit sample:

The pattern, corrected and re-cut, is usually across your base size. Buys you:

Does it fit a human body the way I intended?

Approve this on a person, in a mirror, not in a photo.

03. Pre-production (PP) sample:

The one people skip and the one that matters most.

Made with the actual bulk fabric, the actual trims, on the actual machines, with the actual print or embroidery setup.

Buys you: proof that the thing you approved can be built at scale, exactly as approved. This is the sample you seal.

04. Top-of-production (TOP) sample:

Pulled from the first units off the live bulk line.

Buys you confirmation that the line is producing what the PP sample promised while there is still time to stop it.

What it costs:

A PP sample costs roughly $120-200 and takes 7-10 days.

A 500-piece bulk failure costs $7,000-20,000 and your season.

The trade is not close, and it is never close.

From the floor:

The sample room and the production line are two different environments, and that is the thing brands genuinely do not know.

Sample rooms are staffed by the most senior tailors in the building.

One person often builds an entire garment start to finish, slowly, with care, and they fix small problems as they go without telling anyone they will ease a seam, adjust a notch, or nudge a placement because they want the sample to look right.

The bulk line is the opposite. It is a sequence of operators, each doing one operation, at speed, exactly as instructed. Nobody is quietly fixing anything, because nobody is looking at the whole garment.

So a fit sample proves that a master tailor can build your hoodie.

A PP sample proves that a line can. Those are different claims, and only one of them is the one you are betting 500 units on.

The fix:

Do not compress the ladder; compress everything else.

Build the tech pack earlier, choose the fabric earlier, and start the whole project earlier.

The sampling weeks are load-bearing.

Two hard rules:

Never approve a sample from a photograph, and never go to bulk without a PP sample made in bulk fabric.

Mistake 05 · The safeguard

No golden sample sign-off:

What it looks like:

Bulk arrives. Something is off: the fleece is flatter, the rib has less recovery, and the black is a shade warmer.

You email the factory. The factory replies that it matches what you approved.

And now you are in an argument where neither side has a reference object, which means it is an argument about memory.

You will lose it, and so will they, and the relationship will not survive it.

Why it happens:

Because approval happened over email. Someone wrote “looks great, go ahead” under a photo.

That sentence is not a specification. It is not attached to a garment. It cannot be measured.

What a golden sample actually is:

It is the approved PP sample, physically sealed, signed, and dated by both parties with a duplicate.

Two identical garments, both tagged, both signed.

One stays at the factory as the production reference.

One is couriered to you as your evidence.

From that moment, “approved” stops meaning an email and starts meaning an object.

QC inspects against it. Disputes get resolved against it. It is the contract, in cotton.

What it costs:

Nothing if the run goes well, a courier fee and one extra sample.

Everything, if it does not. Without a sealed reference, the entire cost of any quality dispute is yours, because you cannot prove what you agreed to.

 
From the floor:

Watch what a sealed sample does inside a factory, and you understand why it works.

It hangs at the line, in a bag, with a signed tag. Line supervisors physically hold it against garments coming off the machines.

New operators are shown it on day one.

It converts a subjective standard into an object that a person with no context can use.

That is the entire mechanism; it removes the need for anyone to interpret anything.

Photograph both garments before they are sealed, with the signed tag visible in frame, and keep the images with your PO.

It takes two minutes, and it has settled more disputes than any contract clause we have ever seen.

The fix:

Make sealing a condition of releasing the bulk PO. Two garments, two tags, both signed and dated, and one held at each end. No seal, no PO.

Mistake 06 · The variance trap

Under-specifying color and washes:

What it looks like:

Your black hoodies arrive, and they are three different blacks.

Or the garment-dyed style you loved in sampling comes back across a range from “vintage faded” to “looks like a mistake,” and both extremes are in the same carton.

Why it happens:

Two separate causes that brands tend to blur together.

Because one: color is measured, not named.

“Black” is not a specification. Neither is “off-white,” “sage,” or “washed navy.”

Color has to be communicated as a code, a Pantone TCX reference for textiles, not the TPX paper standard, because dye on cotton and ink on card do not behave the same way.

Then there is the light problem. Two fabrics can match perfectly under the factory’s fluorescent tubes and clearly not match under daylight.

That is metamerism, and it is why color approval is done under a controlled light source, D65 daylight, typically, and not under whatever bulb is above the desk.

And there is the dye-lot problem. Fabric is dyed in batches.

Two batches of the same color, from the same mill, to the same code, will differ slightly.

If your order is split across dye lots and nobody is managing lot continuity, your customers will notice, because they will put two of your hoodies side by side.

Because two washes are variable by design.

Enzyme washes, stone washes, pigment dyes, and acid washes the appeal of every one of them is the irregularity.

That is also the risk. A wash process is not a color; it is a range.

If you approve a single wash sample and say “like this,” you have specified the middle of a range and said nothing about its edges.

What it costs:

Off-shade units are not scrap, which is the small mercy here.

They are markdown.

If 15% of a 500-piece run is visibly off and you discount it 40% to move it, that is roughly $2,000 of margin gone, plus every customer who received the outlier and now thinks that is your quality bar.

From the floor:

The tool that solves the wash problem is boringly simple, and almost nobody asks for it: high and low limits.

Instead of approving one wash sample, you approve three: the lightest result you will accept, the target, and the darkest result you will accept.

All three get sealed. The wash house now has a defined corridor, and QC has an objective test: is this garment inside the corridor or outside it?

Same logic for a dye. Approve the lab dip with a defined tolerance, not just a “yes.”

And ask for the bulk dye lots to be confirmed against the approved dip before cutting because a shade problem caught at the fabric roll is a re-dye, and a shade problem caught at the finished garment is a markdown.

The fix:

Specify color by Pantone TCX code. Approve lab dips under a controlled light source, not on a screen and not under office lighting.

For any wash, seal three samples: low limit, target, and high limit. Require dye-lot continuity across your order, or accept in writing that it will be split.

Mistake 07 · The inventory error

The wrong sizing curve:

What it looks like:

The drop launches. M and L sell out in a week.

You spend the rest of the season sitting on XS and XXL, discounting them, giving them away, and eventually writing them off while turning away the customers who actually wanted the sizes you no longer have.

Why it happens:

Because ordering a flat curve is the path of least resistance.

500 units, five sizes, 100 each. It looks fair. It looks tidy.

It is not how humans are distributed.

Real demand is bell-shaped and skewed by your product and your audience.

An oversized boxy hoodie pulls demand upward; people size up deliberately for the fit.

A slim fit pulls it down.

A womenswear-leaning audience shifts the whole curve.

A flat curve ignores all of it.

A more realistic starting point:

For an oversized streetwear hoodie with a mixed audience, a curve closer to this is a far better first guess than a flat one:

Size Flat curve Modelled curve Units of 500
XS 20% 5% 25
S 20% 15% 75
M 20% 28% 140
L 20% 30% 150
XL 20% 17% 85
XXL — 5% 25
What it costs:

Dead stock is cash frozen, not cash burned, which makes it feel less urgent than it is.

But frozen cash is the thing that kills small brands; it is the money that was supposed to fund the next drop, sitting in a box in the shape of size XS.

From the floor:

A detail worth knowing, because it changes what a bad curve costs: the size ratio does not change your unit price.

Cutting is done from a marker, a layout of every pattern piece across the fabric width, optimized for minimum waste.

The marker is built to your ratio. Whether that ratio is 20/20/20/20/20 or 5/15/28/30/17/5, the fabric consumption and the cutting time are effectively the same.

So getting the curve right is free.

It costs you a conversation and one spreadsheet.

Getting it wrong costs you a season of frozen inventory.

There is no cheaper improvement available to you on this entire page.

The fix:

Never order flat. For a first drop, model the curve from your audience and your fit.

From the second drop onward, use your own sell-through data: sizes that sold out first were under-ordered, and the ones still sitting there are your answer.

Mistake 08 · The last exit

No QC or pre-shipment inspection:

What it looks like:

You pay the balance against a packing list. The cartons ship.

Three weeks later they land, you open box four, and there are skipped stitches on the kangaroo pocket, loose thread ends inside the hood, and a handful of garments with a print sitting visibly off-center.

You email the factory. The goods are in your warehouse, the balance is paid, and every ounce of leverage you had left the building with the freight.

Why it happens:

Because inspection was never a step. It was an assumption: “Obviously they check the garments.” They do.

Against their standard, which you never asked to see, at a level you never agreed to.

What inspection actually means:

Garment QC is a sampling process with an agreed pass/fail threshold, usually expressed as an AQL (Acceptable Quality Limit).

An inspector pulls a statistically determined sample from the finished run, sorts every defect found into critical, major, or minor, and compares the counts against the agreed limits. The lot passes, or it does not.

AQL 2.5 for major defects is a common apparel standard.

The important part is not the number; it is that a number exists, was agreed upon in writing before production, and produces a signed report.

There are three moments to inspect, and they buy different things:

  • Inline / DUPRO during production, at roughly 20-30% complete. This is the only inspection that can still stop a problem. Everything after this can only detect one.
  • Final random inspection once the run is finished and packed. Catches what got through.
  • Pre-shipment before the container is loaded. Your last exit.
What it costs:

A third-party inspection is typically a few hundred dollars.

A 500-piece run with an 8% defect rate is 40 unsellable hoodies at $68 retail, roughly $2,700 of revenue that does not exist, plus the customers who received one before you noticed.

From the floor:

The single most effective thing a small brand can do costs nothing at all: tie the balance payment to a passed inspection report.

Not to shipping. Not to a photo of stacked cartons. To a report with a defect count against an AQL you agreed to before the line started, signed.

It changes the incentive structure of the entire run.

A factory that knows the balance is gated on a report will pull a supervisor onto your line before the report is written, not after.

Every good factory is comfortable with this, because a good factory expects to pass. Any resistance to it is information.

And book the inline inspection, not just the final.

A defect found at 25% complete is a line correction.

The same defect found at 100% complete is 500 garments and an argument.

The fix:

Agree on the AQL in writing before production starts. Book an inline inspection at 20-30% and a final inspection before shipment. Make a passed, signed report on the condition for the balance payment.

Garment QC and AQL inspection: what to book and when

Mistake 09 · The structural one

Ordering trend-heavy with no core:

What it looks like:

Every drop is entirely new. New graphics, new colors, new silhouettes. It sells reasonably well.

And then it is over, and there is nothing to reorder, so drop two starts from zero: new samples, new lab dips, new setup, and new risk.

Two years in, the brand is busy, and it is not compounding.

Why it happens:

Because designing new things is the enjoyable part of running a brand, and reordering the same black hoodie is not. But the black hoodie is the business.

What it costs:

The cost here is invisible, which is why it is the last mistake on this list and the one that does the most long-term damage.

It is not a loss you can point at. It is revenue that never happened.

Think about what a core style earns you that a trend style cannot:

  • Amortized setup. Sampling, pattern development, lab dips, and print setup are largely one-off. Spread across one run, they are a cost. Spread across five reorders, they round to nothing.
  • Better pricing. A repeat style with a known fabric hits fabric minimums comfortably and reorders at volume. Your unit cost falls on exactly the style that sells most.
  • A shorter lead time. A reorder skips proto and fit sampling entirely. The pattern exists, the golden sample is sealed, and the fabric is known.
  • Sell-through data. A style you have run three times has a real size curve, a real return rate, and a real repeat-purchase number. A style you ran once has anecdotes.
  • A reason to come back. Trend products sell to people who saw the drop. Core product sells to people who already own one and want another.
From the floor:

You can see the difference between brands from inside the factory, and it shows up in the sample room.

The brands that survive send us fewer new patterns each year, not more.

Their orders shift from “develop this from scratch” to “run the same block again in a new color, with a new print. The pattern is banked. The fit is proven.

The golden sample is already hanging on the line.

Their cost per unit falls every year, and their lead time shrinks, and none of that is because they negotiated harder.

The brands that stall send a completely new tech pack every season and pay the full development cost of a first-time customer forever.

The fix:

Split each drop deliberately. A workable starting split:

  • 60% core proven styles you reorder. The blank-adjacent hoodie in your best-selling colors. Boring, and it pays for everything else.
  • 30% seasonal core silhouette, new colorway, or new graphic. Low risk, because the pattern and fit are already banked.
  • 10% experimental, genuinely new. This is where the brand gets to be interesting. Keep the quantity small, because most of these will not become core, and the two that do are how you find your next core style.

The system

The Pre-Order Gate:

Nine mistakes, one gate.

This is the checklist we ask brands to run before they release a bulk purchase order: 18 checks across six gates.

Nothing goes to bulk until every box is ticked.

It is deliberately a gate and not a to-do list.

A to-do list is something you get through.

A gate is something you do not pass. Every mistake on this page is a specific failure to close one of these six.

GATE 01

Specification:

Nothing is left to a house standard.

  • Tech pack complete: graded POM table with tolerances, BOM, construction callouts, label, and packaging placement.
  • Every color is named by the Pantone TCX code, not by word.
  • Every trim named by supplier reference, not by description.

GATE 02

Sourcing:

The fabric exists, at your quantity, before the design is locked.

  • Fabric confirmed available: stock-service, or custom dye with the per-color minimum accepted in writing.
  • Fabric consumption per unit calculated and reconciled against the mill minimum.
  • Lead time on fabric confirmed and built into the calendar, not assumed.

GATE 03

Price:

You are comparing specs, not numbers.

  • Identical tech pack quoted by at least three factories.
  • Every quote line itemized: fabric, trims, CMT, finishing, and packing.
  • Any outlier quote interrogated on yarn type, GSM, and compacting before it is accepted.

GATE 04

Sampling:

A line can build it, not just a master tailor.

  • Fit sample approved on a body, in person. Never from a photograph.
  • PP sample made in bulk fabric, bulk trims, on bulk machines, and approved.
  • Golden sample sealed: two garments, signed, dated, one at each end, both photographed.

GATE 05

Variance:

Anything variable has a defined corridor.

  • Lab dips approved under a controlled light source, with tolerance defined.
  • Any wash sealed at three points: low limit, target, and high limit.
  • Dye-lot continuity confirmed across the order or a split accepted in writing.

GATE 06

Acceptance:

You keep leverage until the goods are proven.

  • Size curve modelled from audience and fit or from your own sell-through data. Never flat.
  • AQL agreed in writing before the line starts.
  • Inline inspection booked at 20–30%, final inspection booked pre-shipment, and the balance payment tied to a passed, signed report.

Eighteen checks. Most of them cost nothing but a conversation, and the ones that cost money a PP sample, an inspection are the cheapest insurance in this entire industry.

If you cannot tick all eighteen, you are not ready to release the PO. That is not a delay. That is the gate doing its job.

Questions we get asked before every first order

At Gibben Clothing, 50 pieces per style for most apparel and 100 pieces for headwear.

But the number that constrains you is often the fabric minimum, not the garment minimum.

A custom-dyed color carries a mill minimum measured in hundreds of kilograms, which can translate into hundreds of garments in that one color.

Confirm both before you design.

A tech pack is not overhead you add to a big order; it is the thing that makes a small order survivable.

On 500 units, volume absorbs a mistake. On 50, one wrong spec is the entire margin.

If you are building it yourself, budget time rather than money.

If you are commissioning it, treat it as the cost of not scrapping a run.

No, and this is the single most expensive shortcut on this page.

A fit sample proves that a senior tailor in a sample room can build your hoodie.

A PP sample proves that a production line can, in your actual bulk fabric, with your actual trims, on the actual machines.

Those are different claims. The PP sample is the one you are betting your run on, and it is the one you seal.

Almost certainly split dye lots. Fabric is dyed in batches, and two batches of the same color code will differ slightly; that is normal in textile dyeing, not a defect on its own.

It becomes your problem when nobody manages lot continuity across the order.

Specify that your order runs from a single dye lot where possible, or accept a split in writing and know it is coming.

AQL 2.5 for major defects is a widely used apparel standard, with zero tolerance for critical defects.

The specific number matters less than the fact that a number exists, is agreed upon in writing before production starts, and produces a signed report.

Then make that report a condition of your balance payment.

That one clause does more for your quality outcome than any other sentence in your PO.

Rushing sampling. It is the only mistake on this page that routinely takes an entire run to scrap, and it is the one brands make deliberately, under launch-date pressure, believing they are saving time.

The tech pack is the mistake that causes the most other mistakes. Rushing sampling is the one that costs the most when it lands.

From the HOODIE manual

This is one chapter of The Complete Hoodie Manufacturing Guides

These chapters cover taking a hoodie brand from idea to shipped units, costing, sourcing, tech packs, QC, freight, and launch comparisons. All written from the Sialkot factory floor of Gibben Clothing.

Open The Full Hoodie Guides

On the floor · Sialkot

Written by

Faizan Ahmad

Chief Apparel Technologist & Head of Manufacturing, Gibben Clothing · Sialkot, Pakistan

Faizan leads production at Gibben Clothing, a cut-and-sew streetwear manufacturer in Sialkot, with 8+ years turning raw yarn into retail-ready hoodies, tees, bottoms, jackets, tracksuits, and headwear. He doesn’t just write about clothing; he works the floor, so every guide here is grounded in real fabric behavior, QC standards, and production data from live runs.