Fabric Sourcing for Brands: Mills, Converters, MOQs and What It Costs

There are two completely different fabric markets sharing one vocabulary, and confusing them wastes months. One sells cut lengths to home sewers, sample rooms and small-run makers. The other sells mill runs to brands putting a garment into production against a specification. This page covers the second in detail and tells you honestly when you belong in the first. It sits inside sourcing apparel direct from factories.

Which fabric market you are actually in

Buyer type, the search that usually goes with it, and the supplier that will serve it. Volume figures are per colour.

The honest note first: if you are buying under about 100 metres, buying online by the yard is the correct answer and you will pay a premium for the privilege. Everything below is written for the row where a specification exists.

If you are still deciding what the fabric should be, start with how to write a complete fabric specification and the GSM fabric weight chart before you contact anybody, because a supplier's first question will be what you want and its second will be how much.

Mill, converter and jobber: the three ways brands buy fabric

The three supplier types, with what each one is good and bad at.

Two practical rules. First, never build a repeat carry-over style on jobber fabric — the lot that made your bestseller will not exist when you reorder. Second, converters earn their margin honestly: they aggregate several brands' colours onto one mill run, which is exactly the mechanism that gets you under a minimum you could not meet alone.

What a knitting mill does, and what it does not do

A knitting mill converts yarn into fabric. A circular knitting mill runs cylinder-and-dial machines that produce tubular or open-width knit fabric — single jersey, piqué, 1×1 and 2×2 rib, interlock, French terry and brushed loopback fleece. A flat-knit mill produces fully-fashioned panels for sweaters, which is a different business.

What comes off the machine is greige: undyed, unfinished, and 5–15% away from the gsm you will eventually measure. It then goes to a dyehouse and a finishing line — compacting to control shrinkage, softening, brushing for fleece — before it is fabric you can cut. Many brands assume the knitting mill does all of this. Most do not; fabric moves between three specialist plants, and each hand-off is a place where the specification can be lost.

The exceptions are the vertically integrated groups, and they are worth seeking out. In India, Tirupur in Tamil Nadu runs spinning, knitting, dyeing and garmenting inside one cluster and produces more than half of India's knitwear exports; Coimbatore, an hour away, does the spinning that feeds it. Bangladesh's large knit groups spin, knit, dye and cut-and-sew on one campus. Vertical integration compresses lead time and removes the hand-off risk, and it is a better selection criterion than price on a first order. More on the country picture is in sourcing apparel from India.

Brands very rarely contract cotton yarn suppliers directly. Yarn is bought in tonnes against a count and a staple specification, and the mill or the garment factory is the natural buyer. Your leverage over yarn is exercised through the spec — "100% combed ring-spun cotton, 30s Ne" — not through the purchase order. What that spec buys is covered in combed ring-spun cotton and what it costs.

Fabric MOQs and dye-lot minimums: the real floor on your order

Fabric, not sewing, sets your minimum. A sewing line can physically run 50 pieces. The constraints sit upstream:

• Knitting machine set-up. A circular machine has to be threaded and run in. Short runs waste more in start-up fabric than they produce.
• Dye vessel minimums. A jet, winch or beam dyeing vessel has a minimum load — typically 300–500 kg of fabric per custom colour, roughly 1,000–2,000 tees' worth. Tier-1 Vietnamese mills commonly quote 800–1,200 kg per colour. Below the minimum the liquor ratio is wrong and shade reproducibility fails.
• Piece-dye minimums. 500–1,000 m per colour, about 250–500 garments, at €2–5 per metre.
• General mill fabric MOQ. 500–1,000 m; custom colours 1,500–3,000 m.
• Trim minimums stack separately. Woven labels, custom zippers, buttons and printed elastic each carry their own minimum and often their own colour minimum.

MOQ is a set-up-cost amortisation problem, not a factory-size problem, and two levers actually work: use stock or greige-dyed standard colours that carry no dye-lot minimum, and share one dyed fabric lot across several styles so the colour minimum is met at programme level rather than style level. The full treatment is in dye-lot minimums and MOQ economics, and the choice between yarn, piece and garment dyeing — which changes these minimums completely — is in garment dye, piece dye and yarn dye compared.

Greige versus finished, and how cotton dyed fabric is quoted

Two vocabulary items decide whether you can compare two quotes at all.

Greige is undyed, unfinished fabric straight off the knitting machine. Finished fabric has been dyed, compacted and softened. The two differ by 5–15% in gsm, so a quote for "180 gsm" that turns out to be greige is not the fabric you specified. Write "finished" and a tolerance into every weight you state. A related term is PFD — prepared for dye — which is fabric processed and held undyed, ready for piece or garment dyeing.

Knit cotton dyed fabric is normally quoted per kilogram; woven fabric per metre. Knitting output is measured by weight, so kilos are the natural unit. But there is a second, more expensive ambiguity: mills quote fabric either as finished dyed fabric per kilogram, or as yarn plus a separate conversion charge for knitting, dyeing and finishing. Both are normal and neither is disclosed unless you ask. A $5.50/kg quote and an $8.15/kg quote can describe exactly the same fabric. Cost sheets built on a yarn-basis number come in about a third light against the invoice. Ask which basis, get it written on the quotation, and restate every mill's number on the same basis before comparing.

Worked fabric cost build for a 10,000-piece tee order

Take a 180 gsm single jersey crew tee, one colour, 10,000 pieces, at the canonical finished dyed jersey price of $8.15/kg.

Two checks fall out of this build. At 2,700 kg in one colour you are comfortably above the 300–500 kg dye-vessel minimum and above an 800–1,200 kg tier-1 minimum, so a custom colour is affordable on this order — but split it across five colours at 540 kg each and you are at the edge. And the whole build assumes 1.8 m open width; a narrower width raises your running metres without changing the kilos, which matters only for freight and roll handling, not for cost.

Why buying the fabric yourself usually costs more than letting the factory do it

CMT means you own and supply the fabric and trims and the factory charges only for cutting, making and trimming. FOB or full package means the factory sources materials and quotes a finished garment price. CMT is a commercial model, not a development model — it is not a synonym for cut and sew.

Take the canonical tee cost sheet apart:

Under CMT the factory drops the $2.65 of materials and charges the making side: $0.80 + $0.32 + $0.53 = $1.65. The apparent saving is the $0.85 of overhead and margin the factory was earning on your materials. Now price what you took on to capture it:

• Worse fabric pricing. You are buying 2,700 kg. The factory buys hundreds of tonnes a year from the same mill. If your price comes in 10% worse, that is $0.22 per unit — $2,200 on this order, a quarter of the saving gone immediately.
• Inbound freight and, if the fabric crosses a border, duty on the fabric.
• Yield risk. Marker efficiency on knits runs 80–88%, and the 8–12% wastage allowance is now your exposure rather than the factory's. Come up short on fabric and you are re-dyeing a top-up lot that will not shade-match.
• Working capital, weeks earlier. You fund the fabric at fabric-purchase date instead of at shipment, on a lead time where custom dyeing alone adds 4–8 weeks.
• A commercial argument you cannot win. If the garment fails inspection and the factory says the fabric was at fault, on CMT that is your problem to prove.

CMT genuinely wins in two situations: you own a proprietary or certified fabric the factory cannot buy, or you are consolidating one fabric across several factories. Otherwise the $0.85 is a fee for someone else carrying four risks, and it is usually good value. Definitions for greige, converter, jobber, CMT and dye lot are in the Yarnstick glossary of sourcing terms.

What your fabric specification has to state before anyone can quote

No mill can price a fabric it cannot picture. A quotable specification names, in this order:

1. Fibre content by percentage, descending — "100% cotton", "95% cotton / 5% elastane"
2. Structure — single jersey, piqué, 1×1 rib, interlock, brushed loopback fleece
3. Yarn count in Ne — 20s, 30s, 40s
4. Spinning and preparation — open-end, ring-spun or compact; carded or combed
5. GSM, stated as finished, with a tolerance
6. Dye method and dye class — reactive piece-dyed, pigment garment-dyed, yarn-dyed
7. Finish and performance tolerances — compacted, softened, shrinkage, spirality, colourfastness
8. Width and put-up — open width or tubular, roll length

Every field you leave blank is a field the mill fills with its cheapest option, and it is entitled to. On calendar: budget 1–4 weeks for fabric and trim sourcing and another 4–8 weeks if the colour is custom dyed, inside a 10–16 week first order from PO to FOB for India or Bangladesh, or 15–20 weeks PO to a US distribution centre once ocean transit and clearance are added.

Frequently asked questions