Leading company databases index the brokerages they have already found. We start from 100M+ classified domains — the entire active web — and read each candidate against your exact acquisition thesis, with verbatim evidence behind every call.
Freight brokerage is one of the few B2B service verticals with a public register: every active broker holds federal operating authority. That should make the universe easy to see. It doesn't.
The register tells you who exists, not who fits. Authority records carry a name, a number, and an address — nothing about whether an operation is founder-run or a captive branch, spot-market or contracted, generalist or built around one demanding commodity.
Entry is nearly free — authority, a surety bond, a laptop — so durable, decades-old specialists sit beside single-person resellers in the same register, using the same forty homepage words. A code-based profile cannot tell you which one you are looking at.
A filing is not a description of a business. Working from registrations and code-keyed profiles, a buyer sees:
Brokerage websites are written for shippers, and shippers ask thesis-relevant questions. Read in full, the same universe yields:
Tens of thousands of licensed intermediaries operate below the visibility threshold of profile databases. Their plain, practical websites contain nearly everything an acquirer needs for a first read. Nobody reads them at scale. That is the premise of our method.
We operate several AI platforms and provide large-scale specialized datasets — 300+ enterprise organisations run on our data, including one of Europe's largest telecom operators, adtech and cybersecurity platforms, and a leading airline metasearch.
The first pass is triage: fast LLM reads across every domain our classifier flags as freight-intermediary territory. It answers one question — could this plausibly fit the thesis — and deliberately over-collects, because a false negative here is a target you never see.
The second pass is deep extraction on survivors. Every page that matters gets read: services, about, leadership, coverage, case studies. Each of the 15 signals is filled with a verbatim quote and a source URL, or marked absent. Nothing is inferred.
Evidence of group ownership zeroes the score. No exceptions.
Every company is scored on the same 15-signal framework. In this vertical, five signals do most of the sorting work — and each behaves differently here than in an industrial-services screen.
"Non-asset," "asset-light," and "our fleet of 40 trucks" describe three different businesses with three different risk profiles. We capture the exact phrasing so a buyer with a pure-brokerage thesis never wastes a call on a carrier wearing a brokerage title.
Spot-market language reads very differently from "dedicated lanes," "contract freight," or "managed transportation programs." We log which vocabulary the site actually speaks, with the sentences that prove it — a durability proxy you can audit rather than assume.
Brokerage is a founder's industry: low capital to start, decades to build a book. We record founder and family language only when it is explicit — roughly half of confirmed fits carry this evidence, and brokerage behaves similarly.
Produce and reefer programs, chemical and hazmat freight, flatbed and oversize project cargo, drayage at named ports — specialization is where margins hide. We evidence it from named customers and commodity pages, not keywords; a one-line "we ship everything" claim scores as the generalist it is.
"A division of," "part of the family of companies," an agent-network badge in the footer — brokerage consolidation leaves fingerprints all over websites, and roughly 1 in 10 keyword-perfect candidates turns out to be already group-owned. We exclude them and show you the disqualifying sentence.
Leadership bench depth, hiring posture, branch footprint, digital-commercial maturity, partner ecosystem, activity trajectory — the full framework applies to every candidate that survives triage. The complete signal list, with definitions, is on the method page.
Both approaches have a place. The difference is where the list starts — and whether you can audit why anyone is on it.
| Question a buyer asks | Profile-based databases | Full-web screening |
|---|---|---|
| Starts from every active brokerage site? | Indexes companies previously found | Starts from 100M+ classified domains |
| Separates brokers from fleets and forwarders? | Shared industry codes blur the lines | Reads the model claim on the site itself |
| Sees contracted-lane vs spot language? | Not a database field | Quoted verbatim, with source URL |
| Handles a custom ICP sentence? | Limited to the filters that exist | Your thesis text is the filter; re-runs included |
| Explains every exclusion? | Absences are silent | Documented exclusions, with reasons |
| Finds polished, well-known mid-caps? | Yes — genuinely good at this | Yes, plus the long tail beneath them |
You target established, well-marketed brokerages with clean firmographics, and your thesis maps neatly onto standard filters. No argument from us.
Your thesis lives in the fragmented long tail — founder-led shops, niche lanes, one-sentence ICPs — or you want a coverage audit of what your current tools miss.
Plain language, as specific as you like — "non-asset brokerages with contracted reefer lanes, no institutional owner" is a fine thesis sentence.
Every brokerage-shaped domain in our classified universe gets a first-pass read against it. Broad in, deliberate out.
Survivors get the 15-signal treatment and three weighted scores. Every claim carries its quote and source URL.
Ranked universe, documented exclusions, and custom-ICP re-runs on any subset — included, not upsold.
Composite candidates, anonymized the way our public specimen report is. Client deliverables carry the live quotes and source URLs.
Regional brokerage, produce and refrigerated freight, contracted lanes for named grocery distributors. Second-generation leadership stated on the about page; no group affiliation anywhere on the site; eleven staff named with roles.
Homepage says "supply chain solutions" — the word brokerage appears nowhere above the fold. The services architecture describes classic non-asset freight brokerage with a chemical-freight specialty. Keyword tools file it under consulting; a full read does not.
Everything about the site matches the thesis — until the news page. One sentence disqualifies the company, and that sentence goes in the file so you can verify the exclusion yourself.
Live site, plausible name, national-coverage claims — but three thin pages, no team, no history, no customers. We do not guess our way past missing evidence; the company is flagged as unresolved rather than silently scored.
A vendor that only shows you winners is asking for faith. We deliver documented exclusions and insufficient-evidence flags alongside confirmed fits, because the refusals are what make the inclusions credible. The same discipline governs everything on our standards page.
Full-web screening reads what companies publish. That boundary is the source of its credibility — and of its blind spots. You should know both before you buy.
None of these limits is unique to us; they are properties of public information. The difference is that we state them in the deliverable instead of papering over them, so your team always knows which claims are evidence and which are open questions.
In practice, buyers tell us the flagged unknowns are as useful as the confirmed fits — they turn the first call from a cold pitch into a short list of specific questions.
Deliverables arrive as structured files your team can load into any CRM the same day — one row per company, one column per signal, quotes and URLs attached. The ranked universe becomes your call sheet; the evidence becomes your first-meeting preparation.
Teams running buy-side long lists typically tier by Mandate Fit and work down. Platform teams running an add-on radar re-screen the universe on a cadence and work the deltas.
Freight brokerage theses often sit alongside asset-based logistics or distribution mandates. These pages apply the same method to neighbouring ground.
By reading the model claim on the site rather than trusting an industry code. Brokerages describe arranging capacity through carrier networks; carriers describe fleets; 3PLs describe warehousing and fulfillment alongside transport.
Where a company blends models — an asset-light hybrid, say — we record the exact language and let your thesis decide, instead of forcing a label.
Yes — that is where this approach is strongest. Specialization lives in service pages, commodity pages, and case studies, not in firmographic filters.
Your thesis names the niche in plain language ("reefer brokerages with contracted grocery lanes"), and every match ships with the sentences that prove the specialization.
We capture authority and bond references exactly as the website states them — as claims, clearly attributed. Re-verifying filings against the federal register is a diligence step your team runs on the shortlist, not a screening input.
What we add is everything the register cannot: model, lanes, leadership, independence.
Each engagement runs fresh against your thesis — the screen reads live sites at run time, not cached profiles. Evidence snippets carry source URLs, so anything can be spot-checked the day the file lands.
Annual monitoring re-runs the screen and delivers deltas: new entrants, ownership changes, and dropped candidates with reasons.
No. No website signal supports seller-intent claims, and vendors who make them are guessing — which should worry you about everything else they claim. Our position is on the standards page.
What we do document: founder-associated, long-established, independently positioned businesses with an identifiable decision-maker and limited visible leadership bench — website-visible context, nothing inferred.
Usually a proof project from €4,900: your thesis, run against the relevant slice of the universe, delivered as a ranked file with evidence and documented exclusions.
If the output earns it, the full universe with deep shortlist runs from €9,900, and annual monitoring from €18,000 per thesis. Custom ICP re-runs are included throughout — see pricing.
One email starts it. We send a specimen report the same day — real screening output with every inclusion justified and every exclusion documented — so you can judge the evidence before spending a euro.