Nearly every machined, stamped, or fabricated metal part passes through a finishing shop before it ships — and the shops publish the two facts buyers care about most: their approvals (NADCAP, mil-spec, ISO) and their capacity (tank lengths, oven dimensions, line sizes).
Our specimen run mapped the smallest independent universe of the ten we published: 93 eligible US shops, every classification quoted.
Surface finishing is a chokepoint industry — every fabricated metal part crosses a finishing shop before it ships. The independent universe is small (93 eligible US shops) and actively consolidating, which makes census-grade sourcing non-optional.
NADCAP accreditation, mil-spec approvals, and environmental permits each take years to earn — an incumbent shop is protected by three documented barriers.
Acquired shops keep their names and sites, disclosing ownership only in about-page lines or division references — invisible to keyword screens, readable by ours.
Miss ten shops and you lose a tenth of the market. Solicit five subsidiaries and you burn five of ninety-three credible first impressions. Census accuracy is existential here.
Tank lengths, oven dimensions, crane capacities, substrate lists — finishing sites publish numbers with units because capacity wins quotes. Evidence extraction at its best.
A finishing shop's site speaks two dialects, and the screen reads both — answering a buyer's first two diligence questions before any call is made.
NADCAP accreditations, mil-spec call-outs by number, ISO certifications, customer-specific approvals — captured as exact claim text, never inflated into scopes the site did not state.
Line sizes, tank footage, part-envelope limits, oven dimensions — dimensional evidence that tells a buyer whether a shop can physically fit the work, quoted with units.
93 eligible US independents; 20 shown in the fixed format — 8 top fits, 5 keyword-missed, 5 documented exclusions, 2 flagged — approvals and ownership evidence quoted verbatim.
Open the Industrial Coatings & Surface Finishing specimen →From the 15-signal framework. Scores compose 70/20/10 — Mandate Fit, Outreach Suitability, Transition Context — and documented group ownership zeroes the total.
The gate signal — NADCAP accreditations, mil-spec call-outs, ISO certifications, FDA-adjacent validation language — quoted exactly as published. In aerospace-and-defense theses this is close to binary: without published approvals, a shop belongs to a different thesis.
Process breadth, substrates, and dimensions as stated — which finishes a shop runs, on which metals, at what part sizes. Finishing theses are usually process-specific, and the sites answer in specification language the screen captures verbatim.
Aerospace, defense, medical, automotive, general industrial — evidenced through approvals held, customers named, and case studies. End-market mix drives qualification cycles and pricing power; tensions between approvals and gallery evidence are noted, never hidden.
The division-line trap — acquired shops under legacy branding, disclosed in about pages and division references. Documented ownership zeroes the scores and the quote ships; roughly one keyword-perfect candidate in ten fails this gate.
Finishing shops are family businesses to an unusual degree — roughly half of confirmed fits carry explicit founder or family evidence. Captured strictly as published: founder-associated, long-established, independently positioned, identifiable decision-maker.
The most recent dated content separates active shops from dormant sites still standing after wind-downs. In a 93-shop universe, three dormant entries materially distort the file — we record the latest verifiable activity date so it reflects operating reality.
Founder-led, Indiana; automotive, aerospace, military and defense end markets. “Strategic focus on high-performance surface treatments… coupled with its compliance as a Nadcap accredited supplier.” The accreditation is quoted as claim text with its source page — the approval gate, pre-answered.
Founder-led, California; aerospace, defense, medical, and industrial customers.
Four process families under one roof with compliance credentials captured as stated — process breadth that a consolidation thesis reads as either platform seed or capability tuck-in, and the evidence supports the conversation either way.
“Homepage lacks the obvious category keywords — keyword-driven databases would likely miss or misclassify this company.” Michigan; ISO 9001:2015 quoted; automotive, medical, and aerospace relevance.
Electropolishing is surface finishing by any thesis definition and by almost no homepage's vocabulary — the structural gap full-web screening closes.
Illinois; long-standing QS/ISO certification quoted; customers among fabricators, stampers, and steel mills. Pickling shows up in no finishing keyword list and belongs in most finishing theses.
Across specimens, a fifth or more of confirmed fits are in this camouflaged class — in small universes they are frequently the least-contacted, best-fit companies available.
“A service-oriented leader in the PVD coatings industry” — and, per its own site, part of a global group's high-performance metals division. Quoted, linked, excluded, scores zeroed.
The shops good enough to be acquired are exactly the shops keyword lists keep recommending; the exclusion file is where that error goes to die.
Adjacent vocabulary, wrong business, four extractable signals. Flagged and documented rather than deleted, because showing the noise floor is part of showing the method. Honest triage over-collects by design; honest reporting shows what fell out and why.
A finishing thesis usually spans two or three of these families — the universe file's process columns let you re-cut the 93 by family combination in minutes, and ICP re-runs refine as your thesis sharpens.
A 93-company deliverable earns trust only when the arithmetic above it is visible. The funnel is reproducible on your thesis — not just ours.
Finishing-relevant candidates emerge by classified category plus process vocabulary (anodize, plate, coat, polish, passivate, pickle) — deliberately over-collected, since a missed shop costs over 1% of the market.
15-signal extraction per shop: approvals as exact claim text, process families and capacity from capability pages, end markets, ownership evidence, founder language, activity dates. Exclusions documented with quotes, never deleted.
Your rules may differ: admit Canadian shops, require NADCAP, drop the powder-coating family, add heat treating. Re-drawing the boundary is an included re-run with the same auditable arithmetic above it.
What the screen does, what it stops short of, and where diligence takes over.
NADCAP's own database and your diligence close the loop — one verification call per finalist instead of ninety-three.
No revenue, EBITDA, capacity utilization, or environmental-compliance status beyond what sites state.
A shop with an undated site may be thriving offline — activity trajectory records the latest visible date, never a verdict.
In a vertical that consolidates steadily, the honest argument is for monitoring over one-shot lists.
Sites that cannot support classification ship with a note — in a universe of 93, even that file deserves a human hour.
Everything we decline to claim — standards page.
At this scale the universe file is less a funnel than a chessboard — every company can be known individually.
Approach with their own published credentials and capacity numbers. A note referencing a shop's actual tank line reads as a serious buyer's letter.
Electropolish, passivation, pickling — the least-contacted, least-competed companies in the universe, invisible to keyword screens.
Which groups hold which processes regionally — worth studying before any outreach, refreshed by annual monitoring as the map shifts.
The surface-finishing specimen ships the same day you ask. A paragraph of thesis gets you a scoped proposal for the full universe or your process-specific cut.
Request the specimen report