Fabric Sourcing FAQ: Arvind Denim, Twill Private Label, and Wool Specification Questions

I've been buying apparel fabric for nine years — denim, shirting, twill, knits, wool suiting. Seven documented mistakes, roughly $34,000 in wasted budget, and a few phone calls I'd rather forget. This is the FAQ I now hand to every new buyer on my team. Short answers, real numbers, no marketing.

What has to be in a fabric spec sheet before I ask for a quote?

Fiber content, construction, weight, width, finish, shade reference. That's the floor. Everything else is negotiation.

In 2018, my first year, I sent out a request for “mid-weight cotton twill, navy” and got three quotes that looked comparable. We took one. The fabric arrived at 205 GSM. The pattern had been built around 260 GSM. Same words in the request, three different products. The re-cut cost about $2,800, plus a two-week delay — or rather, closer to three once you count the re-dye queue.

What should have been in that request:

• Fiber content with a tolerance, e.g. 98% cotton / 2% elastane, ±2%
• Construction: yarn count, ends per inch × picks per inch
• Weight in GSM with a tolerance, e.g. 260 GSM ±5%, measured per ISO 3801 or ASTM D3776
• Cuttable width — not just “58 inches”
• Finish: mercerized, sanforized, soft-washed, whatever applies
• Shade reference plus tolerance
• Test methods, not just test names: shrinkage per AATCC 135 or ISO 6330, laundering colorfastness per AATCC 61 or ISO 105-C06, lightfastness per AATCC 16

If you don't name the test method in the spec, you get to argue about the result later. I have lost that argument. Twice.

How do I compare quotes without getting burned?

Compare landed cost per garment, not price per yard.

Here's the thing: a $2.85/yard quote and a $3.15/yard quote are not the same product at different prices. In 2023 we took the cheaper one. Landed cost came out at $3.41 versus $3.22 for the quote we didn't take. The difference wasn't the fabric — it was overage billed above the 3% allowance, a per-color lab dip fee that never appeared on the quote sheet, and a two-week-longer lead time that pushed us into air-freighting part of the order. That air freight was $1,900 — well, $1,940 with the fuel surcharge.

Total cost on a fabric order is roughly: base price + lab dip/setup + overage and short-ship allowance + freight and duty + testing + the cost of the weeks you spent waiting + the risk of a rework. Base price is maybe 70–75% of that number, in my experience.

I'm not saying the low quote is always wrong. At least, that's not been my experience with uncomplicated greige programs on long timelines. For branded apparel with a fixed in-store date, the low quote has cost me money more often than it's saved me.

Worsted or woolen — why does one word break a wool order?

Both are wool. They're processed differently and they don't behave the same way.

• Worsted: fibers are combed to remove short staples, aligned parallel, spun with higher twist. Smoother surface, stronger, holds a crease. This is what most tailored suiting wants.
• Woolen: carded, fibers cross in multiple directions, lower twist. Bulkier, softer, fuzzier. Flannels, coatings, some knitwear.

Micron matters too. Superfine starts around 17.5–19.5 micron, fine wool around 20–24. If your spec says only “wool suiting, navy,” a mill can honestly quote you either one.

In 2021 I spec'd exactly that — “wool suiting, navy, 260 GSM” — and got a competitive quote back. The swatch looked right on my desk under office light. It was woolen, and for a structured jacket with a sharp shoulder the hand was wrong in the first fitting. We paid around $6,400 in re-weave and lost week — mostly the lost week, honestly.

So this is the wool fabric specification guide I use now, and it fits on one page: fiber content and percentages, micron range, worsted or woolen, yarn count, weight in GSM, weave, finish, shrinkage, and the test methods. One more thing — in the US, the Wool Products Labeling Act requires the fiber content on the label to be accurate. “Wool” on a label does not mean 100% wool.

How do I set color tolerance so I stop rejecting bulk?

Industry standard color tolerance is Delta E < 2 for brand-critical colors. A Delta E of 2–4 is noticeable to trained observers; above 4 is visible to most people. (Pantone color matching guidelines)

Denim is a dyed-yarn fabric. Whether it's a vertical mill like Arvind or a smaller weaver, shade depends on indigo dyeing depth, the number of dips, and the finishing route. Batch-to-batch movement is normal. The only question that matters is whether you wrote the acceptable window down before you approved the dip.

In September 2022 our bulk came back at Delta E 3.1 against a lab dip we'd approved at 1.4. Our spec said “reasonable match.” Our customer's QA had a different definition of reasonable. The claim came to $9,200 — no, $9,400; I'm mixing it up with the freight line on the same invoice. 4,500 pieces, all of it technically fine fabric, none of it sellable against the standard we approved.

Lesson: put a number in the contract, and decide in advance who owns lab dip risk. If you approve a dip, you have approved the standard. That's true for Arvind denim fabric and it's true for anyone else.

What does “private label” actually mean on a twill order?

Three terms get used interchangeably and shouldn't be:

• White label — the mill's existing product, your brand on the label.
• Private label — made to your spec, sold under your brand.
• OEM — made to your brand and frequently your pattern, trims, and packaging.

For a twill fabric private label program, settle five things before the first PO: who owns the construction spec, who owns the shade standard, MOQ per color, the overage and short-ship allowance, and the exclusivity window. Get those in writing or you'll negotiate them during a problem instead of before one.

My own worst moment here was small and stupid. We shipped 3,000 woven labels reading 98% cotton / 2% elastane. The third-party test report on the bulk said 97/3. Both were inside our spec tolerance. The label wasn't. $450 in labels, a reprint, and a re-tagging job that ate two days. The composition on the label has to match the tested composition of the actual bulk — not the composition on the PO.

Do polyester fabric manufacturers really differ that much?

Yes, and the difference starts at the yarn, before anything reaches a loom.

Ask for the yarn spec: filament or spun, denier and filament count (150D/48F means something specific), FDY vs DTY vs ATY, luster (bright, semi-dull, full-dull), and whether the yarn is textured. Two quotes can both read “polyester, 150D” and land completely differently on shrinkage, pilling, and hand feel. If a vendor can't tell you which yarn they're running, they're quoting you a hope.

Recycled is its own category. If the spec says rPET, ask for a transaction certificate from a recognized chain-of-custody scheme — not the logo on the website. An unverifiable recycled claim is a liability, not a selling point, and I'd rather lose a spec argument than publish a claim I can't document.

This is where an integrated operation earns its keep. With a mill that runs its own spinning, weaving, and finishing — the Arvind mill setup, for instance — a yarn-level question usually has a documented answer within a day. With a trading house, it may or may not, and you tend to find out during the claim rather than before it.

What timeline should I actually plan for — the one nobody puts in the quote?

Lab dip approval is not the start of production. It's the start of the queue.

The most frustrating part of fabric sourcing: the same slippage happening every season, for reasons everyone already knows. You'd think confirming the lab dip four weeks early would buy buffer — but approval just moves you to the next line. Then finishing, then testing, then documentation, then booking.

Rough numbers from my own log, and your mileage will vary by mill and season: lab dips 1–3 weeks, bulk dyeing and weaving 3–5 weeks, finishing plus testing 1–2 weeks, packing and documentation 3–5 days, freight 2–5 weeks depending on mode. Now subtract factory shutdown windows — the Diwali period, the year-end holidays, and the pre-CNY crunch if any part of your chain touches China.

Two weeks of buffer is the cheapest insurance I buy, and it doesn't show up on any quote. That's the whole argument for total cost thinking in one line: time is a cost line too.

So what's on the pre-check list?

Short version, the one taped above my desk:

• A named test method for every claim in the spec
• Weight with a numeric tolerance, and a cuttable width
• A shade standard with a Delta E number and a named owner of lab dip risk
• Overage and short-ship allowance in the contract
• Label composition matched to the bulk test report
• Landed cost compared, not unit price
• Two weeks of buffer built in
• Any recycled or compliance claim backed by a document

None of that is clever. It's just cheaper than the alternative — about $34,000 cheaper, in my case.