Author Archives: Yang

We are Renrui Textile, a garment factory in Suzhou city of China. So many buyer often sent the E-mail to talk with us about the import duties from China to the USA if they place the bulk order to us what the tariff on their order is going to be, and we’ve learned the honest answer is longer than they want it to be — because the real answer is two numbers stacked on top of each other, and one of them has moved more than once since we started writing quotes this year.

If you check with the freight forwarders and they only send a tidy percentage list but it seems to be useless for you. If you check with us, and you will get the number tied to your actual HS code and your actual production date — because we’re the ones who have to write it on your quote before you send a deposit, not the ones trying to get you to book a container.

When our customer forwarded us a quotation list from other supplier with a single line — “duty: 12%” — we knew exactly what had been left out. The base MFN rate on a cotton hoodie (HS 6110.20.20) is closer to 16.5%, and even that is never the whole bill.

A base MFN (Most Favored Nation) rate, set by whichever HTS subheading your product actually falls under — search by the code, not the product name, because customs doesn’t care what you call it. “Hoodie” and “sweatshirt” can land in different subheadings depending on the knit.

A Section 301 tariff stacked on top — country-specific, separately published, and the layer that’s actually shifted under us mid-quote before.

We will check hts.usitc.gov website by ourselves before each time when we quote hoodie — not because we don’t know our own product’s code by now, but because the rate attached to it has moved before, and printing yesterday’s number on today’s quote is how a factory loses a customer’s trust in one invoice. The Section 301 rate is published and updated by the Office of the United States Trade Representative at ustr.gov — that page, not a blog post, is current the day you read it.

Anchor these two to the real official pages when the post goes live: hts.usitc.gov (HTS lookup) and ustr.gov (Section 301 actions/lists). Do not link a third-party “tariff guide” site as the authority — that undercuts the entire point of this article.

We pulled five tariff guides while writing this, published within weeks of each other. One said 7.5%. Another said a new country-specific rate had just replaced a temporary global tariff, putting the combined number in the low-to-mid teens. One said combined duty on Chinese apparel runs “well above 30% in most cases.” All three were right — for the week they were published, for the HS code they happened to be talking about. None of them were talking about your order.

That’s the actual lesson, and it’s also why we stopped printing a fixed percentage in our own marketing a while back: the rate moves at any time, sometimes inside the gap between your deposit and your ship date. A number we can’t stand behind on the day the container clears isn’t a number worth printing.

Numbers are easier to trust when you can see where they came from, so here’s one built off our own published pricing instead of a hypothetical.

Add duty on top of the FOB total, then add international freight, US customs brokerage, and last-mile delivery, and you get your landed cost per piece — the number that actually decides your margin, not the FOB price on the quote. Two ways this plays out once the container’s by sea.

Route 1 — FOB, you handle customs

We quote it as FOB price for you, and then you need pay the sea freight, duty, and brokerage separately once your goods clear a US port. You keep the pen on which forwarder and which broker you use — and you also keep three separate invoices to reconcile before you actually know your per-piece cost. If your broker files the wrong HS code, you find out after the container’s already sitting in a bonded warehouse, not before.

Route 2 — DDP, one number, no customs handling

We quote delivered duty paid instead: freight, duty, brokerage, and last-mile are already folded into one number before you approve the order. You’re not adding up three invoices to find your real cost, and if this is your first time clearing a container through US customs, you never have to learn how on this order — we’ve already done it on someone else’s. This is the version most private label orders go with, because it turns your landed cost into one number you can put straight into your own retail math.

DDP isn’t automatically cheaper. You’re paying us to hold the customs risk instead of holding it yourself, and if you already have a broker and a freight rate you trust, FOB can win. Ask for both numbers on the same quote. Most suppliers will only hand you one — usually whichever is easier for them to write.

The same two-layer math applies whether you’re pricing a custom sweatpants order or a hoodie run — the HS code changes, the stacking logic doesn’t.

This is the part that made us write this article in the first place. Most sourcing content — ours included, until a customer called us out on it two quotes ago — says “budget roughly X% for duty” and stops there. A range isn’t a number you can put in a spreadsheet and defend to your own buyer.

On your quote, before you pay a deposit, you get:

• The base MFN rate for your HS code
• The Section 301 rate in effect as of your production date, not the date you first asked — that’s the date the rate that actually applies gets locked in
• The combined rate, written as one line — Combined duty: [rate]%, as of [production date] — tied to your order value, not a range
• A DDP quote next to the FOB quote, so you can compare the all-in number instead of doing your own duty math on the back of the invoice

If a sourcing guide handed you a duty percentage without asking your HS code or your production date, that number was a guess wearing a suit. Ours is arithmetic against your actual purchase order.

We have this open in a browser tab every time we build a quote — not a polished internal tool, just a spreadsheet with the same ten rows below. Copy it into your own and run it against any supplier’s FOB number, not only ours.

No email required to use it. If you’d rather hand us your quantity and target garment and let us fill in both the FOB and DDP rows ourselves, that’s what the quote request on our hoodie page is for.

Download the Excel calculator

These are the same pages Renrui Textile checks before writing any US quote.

• U.S. International Trade Commission — HTS lookup: hts.usitc.gov
• Office of the U.S. Trade Representative — Section 301 actions: ustr.gov
• U.S. Customs and Border Protection — import valuation guidance: cbp.gov

The real question is not which filling is "better." It is which construction keeps either one from sliding to the corners after the twentieth wash — and that is the part every blog post about this topic skips, because it is not a filling question. It is a factory question.

We are Renrui Textile, a home textile OEM manufacturer in Suzhou, China. We run glass beads and poly pellets side by side on the same quilting lines, and the choice between them matters less than most people think. What matters more — and what this page will explain with specifics — is the stitch pattern holding them in place.

There is a free spec sheet at the bottom. Excel and PDF, no email required.

You will see “glass beads are 2x denser than poly pellets” repeated everywhere online. We looked for a single source that actually measured it and could not find one — so we will not repeat it as fact. What we can tell you from running both on the same production floor: beads let us build a slimmer panel for the same target weight.

Both materials are available across our weighted blanket OEM line. Both carry OEKO-TEX Standard 100 and GRS 4.0 certified shell fabric options. Neither choice locks you out of a certification requirement.

Both fillings shift if the panel is too big. That is a stitching decision, not a filling decision — and it is the part every comparison article skips because it cannot be judged by reading a spec sheet online. It has to be built.

Baffle-box construction stitches internal fabric walls that form small 3D pockets — the fill can move slightly within its own box but cannot migrate to a neighboring section or pool at one edge. It costs more machine time and more fabric (the internal walls use extra yardage), which is why most DTC brands do not offer it: they buy finished goods off a shared production line and cannot control the construction.

Channel-stitch construction sews straight-line channels instead of boxes. Faster to produce, lower cost, and still keeps the fill from sliding end-to-end — the tradeoff is the fill can settle slightly lengthwise within its channel over time, which shows up as a firmer strip near the bottom edge after months of use.

We run both. Baffle-box for lines positioned on even weight distribution as the selling point; channel-stitch for price-competitive lines where the channel settling is an acceptable tradeoff. Either way, every unit gets weighed before it ships — not sampled, checked — so the spec you approved on the sample is the spec that ships, regardless of which construction you pick. Fill weight is verified within ±3% per unit, in addition to our standard 3-point in-line QC to AQL 2.5.

The widely cited guideline — used by the Sleep Foundation and most manufacturers — is 10% of body weight, with a comfortable range of 5–12%. This is not a hard rule (individual preference varies), but it is a reliable starting point for product development. Below is the full weight-to-body chart, included in the free download at the bottom of this page.

Safety note: Weighted blankets are generally recommended for children aged 3 and older who weigh at least 50 pounds. For therapeutic or medical use, confirm the target fill weight with a healthcare provider. These numbers are product-development guidelines, not medical advice.

Before anything else, read the byline on whatever guide you are reading. Search “how to start a clothing line” and the results are written, almost without exception, by print-on-demand platforms. Their guides are long, genuinely well made, and free — and they all arrive at the same conclusion, which is that you should use print-on-demand. That is not a conspiracy. It is just what happens when the people answering the question get paid based on your answer.

We are Renrui Textile, a garment factory in Suzhou, China. We are on the other side of that trade: we make money when you place a cut-and-sew order, and we make nothing when you choose print-on-demand. So take everything below with that in mind — and notice that we are going to spend the next section telling you exactly when print-on-demand is the right call anyway, because the numbers say so, and pretending otherwise would waste your money and our time.

What follows is what it actually takes to start a clothing line, written from the factory floor: real costs, real minimums, real timelines, and the line items nobody puts in writing.

The figure you will see repeated is that starting a clothing line costs “$2,000 for print-on-demand, or $50,000 and up for cut-and-sew”. That number is not a lie exactly, but it describes a brand launching a full seasonal collection across many styles and colourways, with a designer, a pattern maker and a container.

It is not what a first order costs. Here is the actual arithmetic for a first production run, using our own published FOB prices.

At a $60 retail price, that is the difference between a 40% gross margin and a 75% one. On 100 units it is roughly $1,600 of margin you either keep or hand to a platform. That gap is the entire reason cut-and-sew exists.

So where does $50,000 come from? It comes from doing it at scale, in many styles, all at once. It is not the entry price. The entry price is one style, one or two colours, 100 pieces per colour. At our tee pricing that is somewhere around $90–150 of garments plus a sample. That is the real floor, and almost nobody writing about this will tell you, because almost nobody writing about this is a factory.

Here is the part we lose money on. Print-on-demand is the correct decision, and you should not be talking to a factory yet, if any of these is true:

• You have not proven that anyone wants this. If you do not yet have an audience, a waitlist, or a single pre-order, buying 100 pieces of anything is buying inventory to store in your bedroom. POD lets you find out whether the design sells before you spend money making it.
• The graphic is the product. If your brand’s whole differentiation is the artwork on a standard tee, a standard tee is what you need. The blank is not the point. Cut-and-sew would be paying to customise something your customer is not buying you for.
• You want dozens of designs live at once. Testing thirty graphics is a POD workflow. It is not a manufacturing workflow, and any factory that tells you otherwise is quoting you thirty separate development projects.
• Cash flow cannot survive a deposit. Manufacturing means paying before you sell. If that would break you, POD’s pay-as-you-sell model is not a compromise, it is the correct tool.

We are a factory saying this in writing because it saves everyone a wasted month. Roughly a third of the enquiries we get should be print-on-demand orders, and telling them so at the first email is cheaper for us than sampling something that will never reach bulk.

Where print-on-demand stops working

And then, usually somewhere between the second and the tenth month, three things happen at once.

The margin stops. You cannot discount, you cannot run a bundle, you cannot afford wholesale, and you cannot pay for ads — because there is no room in $28–38 of cost against a $60 retail. POD margin does not improve with volume. It is a fixed ceiling, and once you hit it, the only growth left is more traffic at the same thin margin.

The product stops being yours. You cannot change the fabric weight. You cannot change the fit. You cannot move the shoulder seam, use a heavier rib, line the hood, or put your own woven label in the neck. Every competitor printing on the same blank has an identical garment. When a customer says “the quality feels cheap”, you have no lever to pull.

Reorders become the business. The day a style starts selling repeatedly is the day the maths flips, because a repeat cut-and-sew order needs no new sampling — for us it ships in 10–15 days, and the unit cost drops again at 500 pieces.

The honest summary, if you are deciding how to start a clothing line right now: POD is a testing tool. Cut-and-sew is a business model. Use the first to find out what sells, then switch. The mistake is not starting with POD; the mistake is staying on it after it has told you the answer.

These are our published FOB prices at small-batch volume — 100 pieces per colour — and at bulk. Not a range invented for an article; the same numbers we quote.

On top of the unit price there are exactly three things, and no others:

• A sample fee — the table above. Ours is credited back against your bulk order, which is worth checking anywhere you buy: a factory offering a free sample is making a sales tool, and a factory charging for one is making a garment. Pay for the sample.
• Labels and packaging — $0.35–$0.80 a unit for a woven label, polybag and hangtag. One thing to know before you design a woven label: the loom has a fixed setup, so woven labels have their own minimum, typically 100 per colour. Printed labels go lower. At 100 pieces, plan for that.
• Freight — roughly $2.00–$3.00 a unit by sea at these volumes. Air is faster and several times that.

There are no tooling charges, no setup fees and no plate charges on standard constructions. If a quote contains those, ask what they are for.

So what does the first order really come to?

Take the most common first run we see: one tee style, two colours, 100 pieces each.

100 tees at $1.80–$3.00 is $180–$300. A sample is $8.50–$15, credited back later. Labels and packaging at $0.35–$0.80 adds $35–$80. Sea freight at $2.00–$3.00 adds $200–$300. You are looking at roughly $425–$700 landed, before duty — for a hundred garments that are genuinely yours, in your fabric, with your label in the neck.

That is what it costs to start a clothing line properly, and it is the number the $50,000 figure is hiding. It is not a small amount of money. It is also not a barrier, and it is roughly what a POD brand spends on ads in a fortnight.

Every guide on this topic stops at FOB. Then the goods arrive in the US and the founder discovers that the landed cost is not the quoted cost, and the margin they modelled does not exist.

If you are importing apparel from China to the United States, there are two numbers stacked on top of each other. The MFN base duty, which for most knit apparel sits around 12%. And the current Section 301 rate, which has moved violently and repeatedly, and which no honest factory will print as a fixed percentage on a webpage because it will be wrong by the time you read it.

If you want to see both numbers taken apart on a real order, we did it on a 300-piece hoodie run — FOB against DDP, plus a landed cost calculator you can run on any supplier’s quote: how the two rates stack on an actual invoice.

That is what we do: the duty is quoted as a separate line on the pro-forma, at the rate that actually applies on the day, so the landed cost you model is the landed cost you pay. And if you would rather not touch customs at all on a first import — which for most first-time founders is the right answer — ask for DDP, delivered duties paid, where the goods arrive at your door with duty already inside a single price.

When you compare factory quotes, this is the question that separates them. Most will quote FOB and go quiet about the rest. Ask every one of them: “Will you write the current duty rate into the quote as its own line, and can you ship DDP?”

The single most expensive mistake we watch people make when they start a clothing line is launching a collection. Six styles, four colours each, because that is what a brand looks like on Instagram. That is twenty-four separate production runs, twenty-four fit risks, and a warehouse of the twenty-two that did not sell.

Launch one style. Get it right. Let the market tell you the second one.

Two things about minimums that will shape this decision, and they are not the same thing:

• Fabric dyed to your colour: 100 pieces per colour. Each colour needs its own dye run — that is a physical constraint of the dye vat, not a sales policy. But sizes are mixed freely inside those 100. XS to XXL in one go, with no size multiplier. Ask this question specifically, because a lot of factories quote “100” and then require 100 per size, which is a 300-piece order wearing a 100-piece label.
• Colours are counted separately. Each colour needs its own dye run, so the minimum applies per colour — four colours is four times the minimum, not one. A factory offering you four colours inside a single 100-piece order is either selling you blanks or will renegotiate after your deposit clears.

The practical move for a first run: one colour, all sizes. You find out whether the garment sells before you tie up cash in a colour range.

You do not need to be a designer, and you do not need software. You need to have made the decisions.

A factory quoting your style is reading for three things, in this order: can we source this, can we sew it at the rate we quoted, and what will we be measured against at inspection. Every question your brief does not answer gets answered anyway — by someone in the factory, quickly, according to what is normal for them rather than what was in your head.

In practice you need one of two things:

A reference garment. Post us something you already own and love. A pattern maker can measure it, feel the fabric and read the construction directly, and it will teach a factory more in ten minutes than forty pages will. Say clearly what you want to keep and what you want to change — “this fit, this weight, 3cm longer in the body, heavier rib” is a brief we can act on today. This is the cheapest, highest-value thing a first-time founder can do, and it costs a courier fee.

Or a tech pack. The specification document: flat sketch, bill of materials with fabric weight and composition, measurements in every size with tolerances, construction written as stitch types rather than adjectives, labels and packaging, and your inspection standard. If that sounds like a lot, it is — and it is also the difference between a quote you can rely on and a quote that moves the moment reality arrives. We wrote the full guide separately, with a free clothing tech pack template you can download without giving anyone your email.

You do not need a perfect one to start a conversation. You do need one before you place an order.

From first email to sellable stock is realistically eight to twelve weeks. Here is where that time actually goes, using our own lead times.

Note where the real advantage sits. The first order is slow because everything is being decided for the first time. The second order ships in 10–15 days — which is the entire reason a growing brand stops thinking about unit price and starts thinking about who it manufactures with.

Payment is normally 30% deposit, 70% before shipping, and it should be escrow-protected. Ask.

There is a question underneath this whole process that most founders think but do not ask: you are about to email your complete product design to a factory you have never met, in a country you have never visited. What stops them from making it for someone else?

It deserves a contractual answer rather than a reassuring one. Ask for three things in writing, before you attach a single file:

• An NDA signed before the files move — not after the quote, not with the deposit. A factory that will not sign one before seeing your design has told you something useful for free.
• A no-resale clause in the production contract. This one does more work than the NDA and almost nobody asks for it. It states that your design, patterns and samples are not resold, not listed in the factory’s own catalogue, and not produced for anyone else.
• Ownership of the patterns and tech pack. Your patterns are yours; if you leave, they leave with you. A factory that treats the pattern it graded from your spec as its own property has quietly locked you in, and you will find out on the day you try to move.

We put all three in writing, because it is the question private label buyers ask us more than any other.

Concretely, if you are starting a clothing line from nothing:

• Week 1 — pick one product and one customer. Not a collection. Not “streetwear”. One garment, and a specific person who is already buying something like it and complaining about something specific.
• Week 1 — buy the three garments closest to what you want. Wear them. Wash them. Decide what you would change. This is your brief, and it is worth more than a mood board.
• Week 2 — write the spec. Fabric weight, fit, construction, labels, packaging. Use the tech pack template if you want the structure handed to you.
• Week 2 — email three factories the same brief. Same document, same questions, same day. What comes back tells you more than any directory. Ask each of them: what is your MOQ per colour and is it per size? Is the sample fee credited? Will you write the duty into the quote? Will you sign an NDA before I send files?
• Week 3 — pay for samples. From the two factories that answered best. Yes, both. A second sample costs less than a wrong bulk run.
• Week 4 — sell before you produce. Photograph the sample. Take pre-orders, run a waitlist, post it. Fifty pieces is small enough that pre-orders can genuinely cover it — and if nobody pre-orders a sample they can see, they were not going to buy a hundred of them either.

None of that requires funding, and none of it requires a big order. Starting a clothing line requires asking the questions before production instead of after — and the factories that get annoyed at these questions have told you something useful about themselves, for free.

If you have a reference garment and a rough quantity, you already have enough to get a real quote. It costs nothing, it settles fabric, price tier and timeline in one exchange, and the way a factory answers the questions in this article will tell you more about it than its website will.

We manufacture knitwear and home textiles from our own factory in Suzhou and work with brands from 100 pieces up, which means most people who write to us are placing their first order ever. The categories we run most often are small-batch apparel production, first orders for new brands, hoodies and heavyweight fleece, tees, women’s knitwear and sleepwear and loungewear.

✅ From 100 pcs per colour — sizes mixed freely, no size multiplier
✅ Published FOB pricing — the tables above are what we quote
✅ Duty written into the quote as its own line, or DDP to your door
✅ Sample in 7–10 days, fee credited against your bulk order
✅ NDA before you send files — plus a no-resale clause and pattern ownership in writing
✅ AQL 2.5 — bulk that deviates from your approved sample is remade

⏱️ We reply to every inquiry within 8–12 hours. [email protected] · WhatsApp +86 133 7216 9755

A clothing tech pack is the document that turns a design into instructions a factory can actually cut, sew and inspect from — the flat sketch, the materials, the measurements with tolerances, the construction detail, and the standard the finished goods get checked against. We’re Renrui Textile, a garment factory in Suzhou, China, and this is how work arrives at our door: most weeks at least one tech pack lands that is beautifully designed and still cannot be sewn from — not because the designer did anything wrong, but because it was written for a person, and it is going to be read by a cutting room on the other side of the world.

There is already good material on the internet about what a tech pack is. Almost none of it is written by the people who have to build from one. So this is the other half of the conversation: what a factory actually reads, which fields decide your price, which blank cell quietly hands us a decision that should have been yours — and what gets lost in the gap between your document and a production line that does not share your language, your units, or your assumptions.

There is a free template at the bottom — no email required, no form — and if you fill it in, we will check it free before you commit to a sample. It is the same structure we ask our own clients for.

A tech pack (technical package) is the specification document that turns a design into a manufacturing instruction. It travels with your style through sampling, costing, production and inspection, and it is the reference everyone argues from when something goes wrong.

The definition that matters more, though, is a functional one. A tech pack is the complete set of decisions about your garment. Every decision it does not contain still gets made — it just gets made by someone in the factory, quickly, based on what is normal for them rather than what was in your head.

Which is why the most useful way to build a tech pack is not “what does my design look like” but “what could a stranger get wrong”. A factory quoting your style is reading for three things, in this order: can we source this, can we sew this at the rate we quoted, and what will we be measured against when the goods are inspected. Anything that answers those three questions belongs in the document. Anything that does not is decoration.

Most tech pack templates online come from software companies or content sites — nobody who actually cuts and sews the garment reads them before you send them to a factory. This one does, because we are the factory. Fill it in, send it back, and we will flag anything a production line would misread — free, before you commit to a sample.

If you used an AI tool to generate your tech pack, send it to us first. We have seen what a factory does with the gaps AI-generated specs tend to leave — an empty tolerance column, adjectives in place of construction terms — and it is not cheap to find that out on a sample run instead of a spreadsheet.

1. Cover sheet

Style name, style number, version number, date, category, size range, quantity, target dates. It looks like admin. It is actually version control, and it is the cheapest insurance in the document.

Here is the failure it prevents. You send a tech pack. Two weeks later you change the sleeve length in an email. Three weeks later you change the fabric weight in a WhatsApp message. The factory now holds three versions of the truth, and the pattern room is working from whichever one got printed. Bump the version number every single time anything changes, and re-send the whole document. A file called final_v2_REAL.xlsx is how the wrong spec gets cut.

One field on the cover sheet matters more than all the others: the unit of measure. Declare centimetres or inches once, at the top, and never mix them anywhere else in the pack. More on why below.

2. Flat sketch and callouts

Front view, back view, and detail views for anything that is not obvious — collar, cuff, pocket, hem, closure, artwork placement. Not a fashion illustration: a flat, technical, proportional drawing.

The part people skip is numbering. Number every callout on the sketch — 1, 2, 3 — and then use those same numbers in your bill of materials and your construction sheet. That numbering is what lets a pattern maker connect a line on a drawing to a stitch on a machine. Without it, someone in the factory is matching your paragraph of text to your picture by guesswork, and they will do it at speed.

3. Bill of materials (BOM)

Every material in the garment: body fabric, rib, thread, interlining, labels, hangtag, drawcord, zips, polybag.

And this is where the single most common quoting problem lives. A BOM that says “cotton jersey” is not a specification. It cannot be quoted, it cannot be dyed, and it cannot be knitted. Four fields turn a fabric name into something a mill can actually book:

• Composition — 100% combed cotton, or 95/5 cotton-elastane. The 5% changes the machine, the price and the recovery.
• Weight (GSM) — 180gsm reads as a summer tee, 240gsm reads as a structured piece, 400gsm is a heavyweight hoodie. This is the number that most directly moves your unit cost.
• Yarn count — 30/1 versus 20/1 is the difference between a fine hand and a coarse one at the same weight.
• Finish — bio-wash, enzyme, peached, brushed. This is the “feel” you have in your head, and it is a process, not an adjective.

Give a factory those four and you get a real quote in a day. Give it “soft premium cotton” and you get either a long email or, worse, a confident number that quietly assumes the cheapest interpretation.

4. Measurement spec (POM) — the one with the tolerance column

A points-of-measure sheet lists every dimension of the garment, in every size, with a description of how each one is measured. Chest width 2.5cm below the armhole, laid flat. Body length from HPS (high point shoulder). Sleeve opening, edge to edge.

The description matters as much as the number, because “chest” has at least three plausible meanings and only one of them is yours.

But the column almost nobody fills in — and the reason this section gets its own warning — is tolerance. See below; it is the biggest single point of failure in this document.

5. Construction and stitching

How each seam is joined. Stitch class, seam allowance, stitches per inch, thread, and any reinforcement.

“Clean finish”, “quality stitching” and “nice drape” cannot be sewn. A machine operator does not have a setting for them. What can be sewn: overlock 4-thread, 12 SPI, 1.0cm seam allowance, bartack at the join. If you do not know the stitch names, that is completely fine — send a reference garment and ask the factory to spec it back to you. Any real factory will, and their answer becomes the sheet.

6. Colourways

One row per colour, with Pantone references. Use TCX (the textile cotton library), not TPX or coated chips — those are for print and they will be matched wrong on fabric.

Worth knowing before you plan your colour range: fabric dyed to your colour needs its own dye run, so minimums are counted per colour, not per style. That is why a four-colour range is four times the minimum, not one — and why a first order is almost always cheapest as a single colourway with the full size range inside it. Settle the colour count before you finalise the colourway page; it moves your first order more than any other line in the pack.

7. Labels, trims and packaging

Main label, care label, size label, hangtag, polybag, barcode. Material, size in millimetres, exact placement, and the artwork file.

This is the most-skipped sheet in the entire document, and it is the one that holds a finished order at the door. Fibre content and care symbols are legally required in most markets; the factory prints exactly what you write here and nothing more. Discovering at inspection that your care label is missing a country of origin means re-labelling every unit by hand.

One practical note: woven labels have their own minimum, because the loom has a fixed setup cost — typically 100 per colour. Printed labels go lower. If you are running 100 pieces, that is a real constraint on your first order, and it is better to know it now than after you have designed a woven label.

8. Packing, inspection and sign-off

Folding, pieces per polybag, assortment, carton counts, carton markings — and your inspection standard.

Write the inspection standard into the tech pack, not into an argument after the goods land. Most garment inspection runs on AQL (Acceptable Quality Limit), and the operative word is acceptable: a lot passes with a defined number of defects in it. That is not a loophole, it is how the whole supply chain works — but “passed inspection” does not mean “zero flawed units”, it means the flawed units fell below an agreed threshold.

So agree the threshold. What AQL level, and what counts as a major defect versus a minor one. A loose thread is a minor. A crooked logo is a major. Where that line sits is negotiable, and it belongs in this document, in writing, before production.

Everything above is true for any factory anywhere. This next part is the half that the guides written by design studios do not cover, because they have never had to read a tech pack through a language barrier at eight in the morning with a line waiting.

1. Tolerance — the blank column that decides your bulk run

This is the most expensive empty cell in apparel.

A tolerance is the amount a finished garment is allowed to deviate from your spec and still be correct. Chest width 52cm, tolerance ±1.0cm, means anything from 51 to 53 passes. Fabric relaxes, panels are cut from an 80-layer stack, humidity moves knits — a garment is not a machined part, and some deviation is physics, not negligence.

If you leave the tolerance column blank, the factory applies its own house standard. Not maliciously — it simply has to apply something, and it will apply what is normal for its usual customers. If its usual customers are ordering basic tees at volume, “normal” may be looser than what your fitted style can survive. Then the sample comes back perfect, the bulk comes back drifting, and there is no document that says anyone was wrong.

As a starting point for knits: ±1.0cm on body widths and lengths, ±0.5cm on necks, cuffs and small details. Tighten it where the fit actually depends on it and loosen it where it does not, because tolerance costs money — a tighter tolerance means more rejected panels and slower sewing, and a factory that is quietly holding ±0.3cm across every point of measure is either charging you for it or not doing it.

2. Units — and the mixed-unit tech pack

Designers in the US and UK think in inches. Cutting rooms in China work in centimetres. Both of those are fine, and neither causes a problem on its own.

What causes a problem is a tech pack that uses both — inches on the measurement sheet because that is how the designer thinks, millimetres on the label sheet because that is how the label supplier quotes, and a seam allowance in centimetres because it was copied from a reference document. Every conversion is a chance for a rounding error, and rounding errors compound across a graded size range.

Declare one unit on the cover sheet. Use it everywhere. If you are more comfortable in inches, work in inches — a Chinese factory converts inches every day without difficulty. It is the mixing that hurts.

3. Adjectives do not survive translation

“Clean finish.” “Premium feel.” “Nice drape.” “Not too boxy.”

These are perfectly clear to another native speaker who has held the garment you are imagining. They are close to meaningless by the time they reach a production line — and the honest thing to understand is that they get resolved anyway, silently, by whoever is standing at the machine, according to what “clean” usually means in that building.

The fix is not better English. The fix is replacing adjectives with operations. “Clean finish at the hem” becomes “coverstitch, 2-needle, 2.5cm hem allowance, 12 SPI”. “Premium feel” becomes “220gsm, 30/1 combed, bio-washed”. “Not too boxy” becomes a number in the chest and a number at the hem. Every adjective you convert into a specification is a decision you take back.

4. The reference garment beats the document

If you own a garment whose fit, weight or hand you want to get close to — send it. Physically. It will teach a factory more in ten minutes than forty pages will.

A pattern maker can measure it, feel the fabric, look at how the seams were built, and reverse-engineer a spec that is closer to what you want than your document was, because the garment does not have to survive translation. Measurements travel badly across languages. A physical garment does not travel at all — it just arrives.

This is the single most useful thing a first-time brand can do, and it is free apart from the courier. Send the reference, and mark clearly what you want to keep from it and what you want to change. “This fit, this weight, but 3cm longer in the body and a heavier rib” is a brief a factory can act on immediately.

5. Nobody wrote down what “wrong” means

The last thing that gets lost is the definition of failure.

Your tech pack says what a correct garment is. It very rarely says what an incorrect one is, and those are not the same document. Without an agreed AQL level, an agreed major/minor split and an agreed tolerance, “this is not what I approved” is an opinion — and an opinion loses to an invoice every time.

The three checkpoints worth writing into the pack, in order of how much they save you:

• A PP (pre-production) sample made from bulk fabric, on the production line, approved by you in writing. Your development sample was made by the best tailor in the building, by hand, from whatever roll was around. Your bulk is a different process and often a different dye lot. The PP sample is where you find that out — not after 500 pieces are sewn.
• A golden sample. Two identical copies of the approved sample, signed, dated, sealed. One stays on the factory floor, one stays with you. It converts a future argument from an opinion into a physical fact sitting on the QC table.
• An inline inspection at roughly 20% sewn. Everyone books a final inspection. Very few small brands book an inline one — and inline is the one that saves you, because a systemic error caught at 20% has happened a hundred times and is fixable, while the same error at final inspection has happened five hundred times and your only options are all bad.

There is a question underneath this entire article that most people do not ask out loud, so let us ask it directly: you are about to email your complete product design to a factory you have never met, in a country you have never visited. What stops them from simply making it for someone else?

It is a fair fear and it deserves a real answer rather than reassurance. Three things, and all three should be in writing before you attach a single file:

• An NDA, signed before you send anything. Not after the quote, not with the deposit. Before the files move. A factory that will not sign one before seeing your design has told you something useful for free.
• A no-resale clause in the production contract. This is the one people forget, and it does more work than the NDA. It states plainly that your design, your patterns and your samples are not resold, not listed in the factory’s own catalogue, and not produced for anyone else. An NDA covers secrecy; the no-resale clause covers what happens after your product exists.
• Ownership of the patterns and the tech pack. Your patterns are yours. If you leave, they leave with you. A factory that treats the pattern it graded from your tech pack as its own property has quietly locked you in, and you will discover it on the day you try to move.

We put all three in writing, because private label buyers ask us this question more than any other, and “trust us” is not an answer.

For context, so the timeline in your head is roughly the timeline in ours. This is how it runs at our factory; other factories will differ, but the shape is fairly standard.

Quoting. A complete tech pack gets quoted quickly, because there is nothing to ask about. An incomplete one turns into a round of questions first, and that round is usually where a week goes.

Sampling. Sample in 7–10 days. The sample fee is credited back against the bulk order when you place it, which is worth knowing when you compare against a factory offering a “free sample” — free samples are made to look good, and paid samples are made to be correct.

Bulk. A first order runs roughly 18–26 days once the sample is approved. A repeat order, where the tooling and the spec already exist and nothing needs re-sampling, ships in 10–15 days.

Minimums. 100 pieces per colour on fabric dyed to your colour, because each colour needs its own dye run. Sizes are mixed freely inside those 100 — XS to XXL in one go, with no size multiplier, which is where a lot of factories quietly add one. Working from certified stock fabric we can run 100 pieces split across several colours, which is usually the only realistic path for a genuine first order.

Quality. Inspection at AQL 2.5, and if bulk deviates from the sample you approved, we remake it. Our materials certificates are public and verifiable rather than a logo on a page — OEKO-TEX® Standard 100 SH025 206558 (issued by TESTEX) and GRS 4.0 TE-00110085 (issued by Intertek); take the numbers and check them in the certifier’s own database, which is what you should do with any factory’s certificates. That is a stronger commitment than a defect rate, and it is the one that matters, because “consistent with the sample” is what you were actually promised.

Landed cost. If you are importing to the US, ask for the duty to be written into the quote as its own line — the 12% MFN base plus whatever the current Section 301 rate is on the day. Any factory that will not put the tariff in writing is leaving you to discover it at customs. Door-to-door DDP, with duty included in a single landed price, is worth asking about on a first import.

Whether you work with us or not, those are the questions. A factory’s answers to them will tell you more than its website will.

If you have a filled tech pack — or a reference garment and a rough quantity — the fastest next step is a quote conversation. It costs nothing, it settles fabric, price tier and timeline in one exchange, and how a factory answers the questions in this article will tell you more about it than any brochure.

Renrui Textile manufactures knitwear and home textiles from our own factory in Suzhou, and we work with brands from 100 pieces up, which means most of the people who send us a tech pack are sending their first one. If that is you, the categories we run most often are small-batch apparel production, hoodies and heavyweight fleece, women’s knitwear, cut-and-sew streetwear and sleepwear and loungewear.

✅ Free tech pack review — send it, we flag what a factory would read differently
✅ NDA before you send files — plus a no-resale clause and pattern ownership in writing
✅ From 100 pcs per colour — sizes mixed freely inside that 100, no size multiplier
✅ Sample in 7–10 days — sample fee credited against your bulk order
✅ AQL 2.5 — bulk that deviates from your approved sample is remade

⏱️ We reply to every inquiry within 8–12 hours. [email protected] · WhatsApp +86 133 7216 9755

Almost every cycling brand we talk to opens with the same question, just phrased differently: “Do I put my logo on something that already exists, or do I have you build my kit from scratch?” That is the white-label-versus-in-house decision, and for cycling apparel it carries more weight than it does for a plain tee — because a cycling kit is a technical garment, and the two production models handle “technical” very differently.

As a cycling clothing manufacturer that runs in-house cut-and-sew production, we have an obvious stake in this. But the honest answer is that both models are right — for different brands at different stages. Here is how we explain the trade-off to the cyclists and founders who ask.

White label means the garment already exists. A factory has pre-developed a jersey, a bib short, a jacket — pattern graded, fit-tested, sewn, inspected — and sells that same base to many brands. You choose from what is on the shelf, add your logo and sublimation print, pick your labels, and you are in market.

The appeal is speed and low risk. You skip pattern development, fabric trials and sample rounds, so entry cost is lighter and lead time is shorter. For a new club kit, an event one-off, or a founder testing whether anyone will buy the brand at all, that is often exactly the right call. You are buying a known-good product and putting your name on it.

The limit is that the garment is not yours. The fit block, the chamois, the fabric weight, the panel layout — those were decided for everyone, not for your rider. If your differentiator is “a better-fitting bib” or “a specific aero cut,” white label cannot deliver it, because you do not control the thing that would make it true.

In-house (full OEM cut and sew) means the garment is built to your spec. You bring a tech pack or a reference kit and a fit intent; the factory drafts or adapts the pattern, sources the fabric to your target weight and stretch, sews a proto sample, and you approve the real thing before bulk. Every panel, seam and trim is a decision you get to make.

That control is what costs more — in MOQ, in sample rounds, in lead time. You are paying for development, not just production. The pay-off is a kit that is actually yours: a fit block competitors cannot buy off a shelf, a chamois matched to your ride time, an aero panel layout you specced. Once a brand has proven demand, this is usually where it moves.

With a cotton tee, white label versus custom is mostly about branding and fit preference. With cycling apparel it is about whether the garment performs, because every piece is doing a technical job:

• Jerseys have to move heat and sweat out fast while holding shape at speed — that is fabric choice and panel construction, not just a print.
• Bib shorts live or die on the chamois and the balance of compression versus support. A generic pad under a three-hour ride is where returns come from.
• Skinsuits are about drag: seam placement and fabric tension either reduce it or they don’t, and a shelf product was not optimised for your rider’s position.
• Winter kit needs insulation without bulk — a construction problem you can only solve if you control the construction.

This is the real reason the decision is harder for cycling than for casualwear: the parts of the garment that white label freezes are exactly the parts that determine whether a serious cyclist keeps wearing it. So the question is less “which is cheaper” and more “does my brand’s promise live in a part of the garment I need to control?”

Match the model to your stage, not to what sounds more premium. If you are testing whether the brand sells, or fulfilling a one-off club or event order, start white label — prove demand before you pay for development. If riders already buy from you and your growth depends on a kit that fits or performs better than the shelf option, move to in-house so the differentiator is real.

Plenty of brands run both: white label for quick club and event runs, in-house for the flagship race kit. What matters is being honest about where your brand’s promise actually lives, because that is where choosing a cycling clothing manufacturer with in-house control either pays off or doesn’t. If it lives in a logo, white label is enough. If it lives in the fit, the chamois or the aero, you need to own the build — which is what our activewear manufacturer cut-and-sew program is for, and it sits inside our wider low MOQ clothing manufacturer service from 100 pieces per design.

Search “is bamboo fabric toxic” and you will find Reddit threads full of alarm and very few straight answers. As a factory that produces bamboo sleepwear for brands, we have an obvious interest here — which is exactly why this article separates what is documented from what is marketing, and shows you how to verify every claim yourself, including ours.

The one-sentence answer: the finished bamboo viscose garment you wear is regenerated cellulose and is not toxic; the legitimate concerns are about the manufacturing process and about brands overstating what “bamboo” means. Both concerns have a verification path. Here it is.

Nearly all “bamboo” sleepwear is bamboo viscose (also labeled rayon from bamboo): bamboo pulp dissolved chemically and regenerated into fiber. It is chemically the same fiber family as any viscose — soft, breathable, drapey. It is not the raw plant woven into cloth, and mechanically-processed “bamboo linen” is rare and rough.

This distinction is regulated. The US FTC requires the fiber to be labeled “rayon from bamboo” or “bamboo viscose” — brands have been fined for labeling products simply “bamboo” and implying the natural stalk’s properties transfer to the fiber. When a label says 95% bamboo viscose / 5% spandex, that is the honest, compliant description. When marketing says “made of pure bamboo,” someone is either careless or hoping you won’t ask.

Viscose production uses carbon disulfide as a solvent. Historically, poorly-controlled viscose plants exposed workers to it, and that history is the true core of the toxicity discussion. Two things follow from it:

First: it is a process issue, not a product issue. Carbon disulfide is volatile — it does not ride along in the finished fiber. By the time fabric is spun, dyed, washed and sewn, the question “is there solvent in my pajamas” is answered by lab testing, which is exactly what OEKO-TEX exists to do (next section).

Second: the process concern is real and worth voting on with your wallet. Responsible mills run closed-loop chemical recovery. Buyers who want the strictest version of this choose lyocell-process fibers, where the solvent (NMMO) is captured at 99%+ rates. If a supplier cannot say anything about their fiber source’s process, that silence is information.

OEKO-TEX STANDARD 100 is a finished-product test, not a marketing badge. A certified fabric has been laboratory-tested against a list of several hundred regulated and suspect substances — formaldehyde, extractable heavy metals, pesticides, chlorinated phenols, banned azo dyes, and residual chemistry among them — with limits stricter for products in Class I (babies) and Class II (direct skin contact, which is where sleepwear sits).

Three things it does not do, so you can weigh it honestly: it does not certify the environmental footprint of fiber production (that is the process issue above), it does not verify recycled content (that is GRS territory), and it does not attach to a brand — it attaches to specific certified material, which is why the certificate number matters more than the logo.

Any real certificate can be checked at the issuer’s public database — for OEKO-TEX, the Label Check tool on oeko-tex.com; for GRS scope certificates, the Textile Exchange database. Paste the number, confirm the holder name and validity window match the supplier you are talking to. A supplier who sends a blurry logo image instead of a checkable number is failing the cheapest test in sourcing.

Practice on ours: our OEKO-TEX STANDARD 100 certificate SH025 206558 (issued by TESTEX) and our GRS 4.0 scope certificate TE-00110085 (issued by Intertek). Every fabric we quote for a bamboo program ships under valid certification, and we include the documents with every order — a practice any competent bamboo pajama manufacturer should follow as standard, not as a favor.

Sourcing Bamboo Sleepwear With the Paper Trail Included

If you are building a bamboo program, work with a pajama manufacturer that treats certification as a document you verify, not a logo you trust. Programs start at 100 pieces per color per style; certified bamboo jersey runs $4.00–$5.50 FOB per set at entry volume.

✅ OEKO-TEX® Standard 100 — SH025 206558 (TESTEX), verify it yourself
✅ GRS 4.0 — TE-00110085 (Intertek)
✅ Low MOQ — from 100 pcs per colour / design
✅ Free fabric swatch pack on request

⏱️ We reply to every inquiry within 8–12 hours.