The empty shelf above Supima.

There is no prior Supima work to continue. The memory corpus, the product board (11 families, 54 products) and the supply-research corpus (44 sources — logistics, tariffs, margins, lithium, cobalt, nickel) hold nothing textile. The design-to-market engine’s live pilot is a sonic massage hairbrush.

The sovereign index is blind here too: a supima query returns three results, none of which contain the word. Every fact below comes from open-web research, not from the estate.

“From the first bit up” means the fibre-to-finished-good ladder. Each rung either exists in the US at commercial scale or it does not. This is the constraint that decides everything downstream.

Two domestic terry mills closed in 2024. One of them is complete, six years old, and for sale — and the seller intends to ship it overseas.

Bleakie’s stated preference is “a buyer from India or Pakistan,” and his thesis for making it work is adding 24 more looms running imported yarn under a Made-in-USA label.

That is a commodity thesis, and it is exactly why the mill died. It does not describe a $70 Supima towel — which clears 74% gross margin at $18.15 landed, with 54% cost headroom before it reaches the category floor.

The asset is priced for a business that cannot work in America. The premium position has entirely different unit economics. Demand exists and is unserved; supply exists, sits idle, and is for sale. The gap between them is capital and an operator — which is precisely where state and federal partnership has leverage. Virginia steered 400+ jobs into that plant once, and Bristol lost them.

Two axes matter: fibre provenance and where it is made. Almost everyone premium is imported Turkish or Egyptian. The Supima-and-domestic corner has two occupants, both small.

Read the table as a gap map. Nobody is selling a plush, high-GSM, all-Supima, US-woven terry towel at scale. Authenticity50 is closest and is running out of stock rather than expanding. Onsen holds waffle, not terry. Everyone above $50 is imported.

The cutesy Asian segment is two distinct waves that want opposite things. Conflating them is the main way to get this wrong.

That second panel is the opening. Cotton is native to the cutesy silhouette — ribbed bralettes, gingham, bows, soft cups — but the brands serving it compete on print, not on hand-feel. A Supima-cotton cutesy line is differentiated on the one axis the incumbents have conceded, and LA cut-and-sew can produce it at 50–300 units per style with no mill relationship.

Every US towel brand competes on GSM because it is the only number a shopper sees. GSM measures mass, not performance — a heavy towel can be slow to absorb, and a reviewer measuring Brooklinen’s published 770 got 740. There is no published, falsifiable quality standard in the American market.

Imabari has had one for years. A 1 cm² swatch is cut from every production batch and floated, unwashed, on soft water. If it does not begin to sink within five seconds, it is not an Imabari towel. Four criteria gate the mark: 100% cotton, absorbency, durability, safety — plus colourfastness, strength and dimensional stability in the lab.

Adopt a published absorbency standard, run it on every batch, print the result, and run it on competitors too. It costs almost nothing and converts an unfalsifiable “luxury” claim into a measured one — on an axis where a 590 GSM towel priced at $60 cannot follow.

Supima costs 30–60% more per pound than upland cotton — roughly $4.00/kg in 2025, and falling: prices slid 6.76% from November 2024 through October 2025 as ELS demand cooled. That premium sounds like a margin problem. At retail it is not, because fibre is a rounding error in a finished towel.

Two percent of retail buys the strongest fibre claim in the category and a domestic-origin story no Turkish or Egyptian competitor can answer. Against a category where median DTC gross margin is 57%, premium bedding brands hold 60%+ on DTC and 30–35% on wholesale, and healthy operators run contribution margin above 35% with LTV:CAC between 3:1 and 5:1 — the fibre is not what threatens the margin.

Authenticity50 has already found the cost lever: Supima in the loops, commodity American cotton in the ground weave. The pile is what you touch and what absorbs; the base is structure. All-Supima is a cleaner claim, Supima-pile is a better cost structure, and the difference is invisible in use.

Towels and lingerie look like one business and behave like two. The asymmetry should drive sequencing.

The obvious sequence is intimates first. It de-risks nothing about towels, but it produces revenue, audience and fit data while the mill conversation runs on a slower clock. Starting both on day one is fine; starting towels alone means a long silent period gated on one phone call.

1. Confirm the Georgia weaver. Identify the mill behind Authenticity50 and establish whether it takes contract terry at emerging-brand volume, at what MOQ and lead time. Everything on the towel side is downstream of this answer.
2. Open the Supima license. Fees were removed and the programme moved to per-bale funding on the end user, with digital traceability through SupimaAqre. Low cost, and the traceability is itself collateral.
3. Write the quality standard before the product. Define the absorbency and durability test the brand will publish, source the lab, and run it on the competitor set first — that baseline is the launch asset.
4. Sample the intimates line at three LA shops. Cotton-forward silhouettes, domestic trims only, 50 units per style. Fit and hand-feel data in weeks, not quarters.
5. Seed the crawler with the textile domain. The index cannot see this category. Fixing that turns competitor tracking from a manual session into a standing instrument — pricing, GSM claims, review mining.

Section 06 argued the fibre premium is trivial at retail. That argument is now priced through the sovereign cost engine rather than asserted. Duty inputs are the real ones: 9.1% MFN on HTS 6302.60.0020, 7.5% Section 301 (List 4A) for China, plus the 10% Section 122. Run at 5,000 units, $70 retail.

Domestic clears the 60% premium-bedding floor with $9.80 per unit of headroom — a 54% cost-overrun cushion on the conversion estimate. Margin is not what stops these being made in America. Mill access is.

⚠The FOB figures are declared assumptions, not quotes; everything downstream of them — duty, MPF, HMF, break-even — is computed. Per-corridor FOB is the one input a real mill conversation replaces.

The jobs and security case

Priced separately, because border cost knows nothing about jobs. Modelled on the 180 jobs 1888 Mills eliminated in Griffin, Georgia, at a $52,000 wage with 40% benefits against a $230,000 median home.

The supply risk came out identical — 11 either way. Domesticating onto a single mill buys no risk reduction: the origin advantage is exactly cancelled by the single-source flag on terry weaving. That is the post-1888-Mills situation stated numerically, and it is the strongest argument for treating the second weaver as part of the plan rather than a later problem. Reshoring onto one supplier is not de-risking.

So the trade is legible: 168 jobs and $12.2M a year in local compensation, for 9.4 margin points — with the margin still comfortably above the category floor.

Nishi · category entry brief · built on nishi-ds v2 · published via nx_site_publish · verified from the WAN vantage