Textile note

Your Fiber Problem Isn't Price — It's That Nobody Owns the Spec

2026-10-08 · Sahana Iyer

The Order That Cost Me a Month

In March 2024 I had 2,900 cones of recycled sewing thread sitting in a bonded warehouse because of a certificate mismatch. (Should mention: the PO said 3,000. I stopped chasing the difference after the first week.)

The thread was fine. The paperwork wasn't. The batch number on the invoice didn't match the batch number on the recycled-content certificate, and the supplier's proposed fix was to send a new certificate — which meant waiting on the mill, which meant waiting on the fiber producer behind the mill. Three companies, each one politely explaining that this wasn't really their part of the chain.

Landed cost on that order was around $14,000. The unscheduled cost — demurrage, a partial re-buy from a second source, and maybe 30 hours between me and our sourcing lead — came to something like $4,100. Maybe $4,500 if you count the label reprints. I'm not going to pretend I tracked it to the dollar.

For about two weeks I was sure the problem was price. It isn't. Price is just the easiest thing to argue about in a conference room.

Most Quote Comparisons Happen at the Wrong Layer

I run indirect purchasing for a company of about 300 people — thread, trim, consumables, some packaging. (Should mention: my desk started with office supplies. The thread came later, after a reorg in 2022.)

When I started, my vendor comparison was unit price. Three quotes, lowest number wins. That worked fine for copy paper and it failed badly for anything that ended up inside a finished product.

Total cost of ownership (i.e., everything that shows up after the quote gets accepted) on a fiber or yarn program usually includes:

  • The fiber or yarn price itself
  • Freight, duties, and the packaging format you actually receive it in
  • Lab testing and re-testing for spec and claim verification
  • Line speed loss or downtime when the yarn runs differently than the sample did
  • Rework, re-labelling, and markdowns when a shade or a claim fails
  • Your own time, which nobody puts in the spreadsheet

That last one is what got me. I once spent six weeks managing a claim against a supplier over a nylon yarn lot that was, functionally, acceptable. The dispute was worth maybe $2,800. Six weeks of my time is worth a lot more than $2,800, and I didn't have six weeks to spend. The conventional wisdom is that a diligent buyer fights every claim. My experience across 200-plus POs says the opposite — sometimes the fastest path is to eat it, change suppliers, and move on. (Maybe 180 POs. I'd have to check the system.)

The supplier in that case was "flexible." What I mean is they'd negotiate on price and on absolutely nothing else.

The Spec Sheet Erases the Thing You're Actually Buying

Here's the part that took me longest to understand, and the part most tech packs hide. When a spec line says "spandex, 40 denier" or "viscose, 30s count," that's a class, not a product. It tells you the fiber family and the dimension. It does not tell you which producer made it, under what process, or who is standing behind that description when an auditor asks.

Two real options exist, and they are not the same purchase.

Branded fiber. You're buying a named product from a named producer — Hyosung TNC's creora spandex being the obvious example in elastane, where the fiber arrives with published technical data, a traceability chain, and a producer who will generally get on a call when a mill's process goes sideways. You pay for the name, and what you get in return is someone accountable. Evaluating a nylon yarn manufacturer works the same way: the useful question isn't "what's the denier," it's "whose fiber is this and who signs the test data."

OEM / private label. You buy fiber or yarn produced to your specification and sold under your label or a neutral one. You control the positioning and often the margin. What you also take on is the claim. If a recycled-content percentage is challenged, or a pulp-origin question comes up on the viscose side, the brand on the label is the one answering the letter.

Neither option is wrong. Plenty of programs run generic viscose or commodity nylon for years and it's the correct call, because the end use never tests the difference. The mistake is choosing one without noticing that you chose.

The "branded fiber is just marketing" thinking comes from an era when the only visible difference was the price tag. That's changed. As I understand it, the wood-pulp side of viscose supply chains now sits inside the EU Deforestation Regulation, which is why pulp-origin documentation moved from nice-to-have to required for a lot of buyers. Recycled content is similar territory — GRS and RCS both run through Textile Exchange, and both lean on transaction certificates that have to match your shipment, batch for batch. I'm not a compliance officer, but I've learned that certification schemes are chain-of-custody systems, not stickers.

Nobody Owns the Claim

Here's the thing I didn't see for the first three years.

The failure almost never happens because someone bought a bad fiber. It happens because the sentence "this product contains 50% recycled content" doesn't have an owner.

Follow it through a normal organization. The designer writes "spandex, 40D" in the tech pack because that's the level of detail that matters for fit. The mill buys fiber against a price target because that's what protects their margin. The brand's compliance team finds out at the lab, or worse, at a customer audit, that the certificate and the shipment don't line up. By then the goods are sewn.

Every party did their job. Nobody did the job.

I've watched a spec get loosened to save three cents a yard, and I've watched that same three cents turn into a re-label because the recycled percentage came in under the marketing claim. The fiber was fine in both cases. The ownership was missing.

What It Actually Costs

The direct costs are the ones that show up in a spreadsheet, and they're the smaller half.

Direct: re-testing, re-labelling, expedited freight when a replacement shipment has to fly instead of sail, and the finished-goods markdown if the window closes. On my March order that was roughly $4,100, hedged.

Indirect, and worse: the mill loses confidence in your spec and starts building in its own buffers, which means longer lead times and higher quotes forever after. Your compliance team starts sampling more, which means more lab invoices. And you personally start avoiding a category because it burned you once, which is how companies end up with a single-source dependency they didn't plan.

If you're sourcing recycled sewing thread wholesale at any volume, the certificate chain is not an administrative detail you handle after the PO. It's the product. Everything else — the tensile strength, the colorfastness, the cone put-up — is checkable at the dock. The claim is the part that can hold your container hostage.

What I Do Now

Three rules, and I'm not going to pretend I arrived at them elegantly.

One: name the product, not the class. If the spec says "spandex 40D," that's a family. If it says the fiber by name, with the producer identified, then the mill can't quietly substitute and the lab can tell you what you actually got.

Two: decide branded vs OEM/private label on who can carry the claim — before you compare unit price. On a program where the recycled claim is customer-facing, the fiber that comes with a traceable chain is probably the cheaper option, even at a higher quote. On a program where nobody will ever audit it, private label is fine. That's a judgment call, and it's yours to make.

Three: verify the certificate before the PO, not after. Ask for the transaction certificate with the batch number on it, in advance, and check that the batch matches what's shipping. That single step would have saved me the $4,100.

I've accepted a lot of cheap fiber quotes over the years. I don't remember any of them. The one I remember is the order where the paperwork didn't match the thread.