Your fund has one thesis and a finite clock. We screen the entire active web against that thesis — every eligible company in the niche, each with verbatim evidence, none recycled from another searcher's list.
Most funds budget roughly twenty-four months. Subtract diligence and closing, and the window for finding your company is shorter than it looks on day one.
In the niches we have screened end-to-end, the entire eligible universe of independent US companies runs from 702 shops in precision machining down to 93 in surface finishing. Small numbers change strategy.
The uncomfortable arithmetic: a searcher who mails 60% of a niche is not running a 60% search. The best-fit companies cluster in the unmailed tail precisely because harder-to-find businesses get less competing mail. Coverage is not a vanity metric; it is where proprietary deals come from.
A profile-based database export, filtered by industry code and headcount — often the same export the last three searchers in the niche pulled.
We start from 100M+ classified domains — effectively the whole active web — and read each candidate site against your written thesis.
A fifth or more of confirmed fits in our specimen runs never used the category's obvious homepage keywords. Those companies do not exist in keyword-driven exports — yet they answer cold outreach like anyone else. Find them first, and you are the only buyer in the room.
The two-pass architecture spends cheap analysis broadly, then expensive analysis narrowly.
Your acquisition criteria, written down and agreed — subsector, geography, model.
From 100M+ domains to the candidate set — one industrial category alone held 367,478.
A 25,000-domain US triage resolved to ~17,300 live operating companies.
Full-site reading against 15 signals, quoting the site verbatim for each.
Mandate Fit 70%, Outreach Suitability 20%, Transition Context 10%.
These figures come from screening work we have actually delivered — the same pipeline your thesis would run through.
The category cut is the honest denominator behind every count we quote. When we say a niche holds 254 eligible independent calibration and testing companies, that number survives an IC's "compared to what?" — see the specimen report.
Our framework extracts fifteen signals per company, every one grounded in a quote from the company's own site. For a searcher planning years of personal involvement in whatever they buy, these six carry the most weight.
Explicit founder or family language on the site — "second generation", "family-owned" — never inferred. Roughly half of confirmed industrial fits carry it, and it shapes every first conversation.
How many principals the site actually names. A one-name bench means the operator you back is replacing the whole visible leadership — priced into your plan, or a surprise later.
Stated founding year and independence language. Long-established shops with continued independence are the classic search profile, and the site usually says so in its own words.
Signs a company already belongs to a consolidator. One in ten keyword-perfect candidates fails here; the score zeroes out and the exclusion is documented with the quote that proves it.
Service contracts, maintenance agreements, scheduled programs, consumables — the visible language of repeat revenue that a first-time CEO can stand on while learning the business.
Explicit claims only — ISO 9001, AS9100, ITAR, UL 508A, ISO/IEC 17025 — captured as exact claim text. Certifications are moats, and they are checkable in diligence.
From 702 eligible independent US shops, the specimen shortlist documents 8 top fits, 5 keyword-missed fits, 5 documented exclusions and 2 insufficient-evidence flags. Four rows, anonymized:
Massachusetts, 58,000 sq ft facility. AS9100D, ITAR and CMMC Level 2 — each captured as exact claim text.
Ohio, ISO 13485. Operating for more than eighty years with visible generational continuity on its history page.
CNC machining for defense, aerospace and medical-device customers — but the homepage never uses the category's obvious keywords.
Keyword-driven databases would likely miss it. Full-site reading did not.
A keyword-perfect contract machining company — excluded as already part of a group, with the acquisition statement quoted in the deliverable.
That is one letter you never send, and one IC question you can answer before it is asked.
| What a searcher needs | Broker / intermediary lists | Profile-based databases | Full-web screening |
|---|---|---|---|
| Covers companies nobody indexed yet | Only marketed deals | Only the indexed fraction | Starts from the whole active web |
| Finds fits without obvious keywords | Not applicable | Keyword and code driven | A fifth+ of fits surfaced this way |
| Flags group-owned companies upfront | Rarely | Ownership data lags | 1-in-10 exclusion rate, documented |
| Proprietary to your search | Shown to every buyer | Same export for every subscriber | Run against your thesis alone |
| Evidence you can re-verify | Seller-provided | Profile fields, source unclear | Verbatim quotes + source URLs |
Brokered processes and databases both have their place — we use category data ourselves. The claim is narrower: for proprietary outreach in a defined niche, the census wins on coverage and evidence.
Ten industrial-services subverticals from one full US run, each figure the count of eligible independent companies after triage, deep extraction and group-ownership exclusions. If your thesis touches one of these, part of your denominator already exists.
A tool you trust on the way in should be honest about its edges. Ours has four.
The deliverable is a working file, not a PDF to admire. Scores sort your outreach tiers; evidence snippets become the first line of a letter that proves you did the reading.
Owners answer specifics. "I noticed you've held UL 508A certification since 2011" outperforms any template — and the quote is sitting in the row next to the score.
A Proof project from €4,900 screens one niche end-to-end. The full universe with deep shortlist starts at €9,900; annual monitoring at €18,000 per thesis.
Send us the thesis you're searching on. We'll scope the niche, quote the timeline, and show you a specimen of exactly what lands in your inbox.
We publish what we refuse to claim — no intent flags, no financial guesses, no owner profiling. Read our standards.