Every buy-side mandate review contains the same trap question: “how do we know this list is complete?” If the honest answer is “we exported the same database you could license yourself”, the advisor is selling formatting.
We build long lists from the entire classified web, with a quoted evidence trail for every inclusion and a documented reason for every notable exclusion.
Buy-side advisory has an awkward secret: the deliverable that anchors the entire mandate — the long list — is usually the least differentiated thing the advisor produces.
The process letters are crafted, the approach sequencing is thoughtful, the negotiation is genuinely expert.
But the list itself, the thing that determines what all that expertise gets applied to, typically began as a filtered export from the same two or three databases the client's own corp dev team licenses, topped up with the partners' personal knowledge of the sector.
Sophisticated clients have noticed. Corp dev teams now routinely run the advisor's list against their own tools during the first review, and the uncomfortable meetings are the ones where the overlap is total —.
because then the advisor's coverage claim reduces to “we know the same companies you do, formatted better.”.
The mandates that renew are the ones where the advisor put companies on the table the client had never heard of, in the client's own vertical.
That is a high bar precisely because the client is expert; clearing it requires a source pool the client does not have.
There is also a quieter cost inside the advisory firm. Long-list construction is analyst work — weeks of it.
Juniors google candidate after candidate, skim sites, guess at ownership, and paste findings of uneven quality into a tracker no one fully trusts.
The output is expensive, slow, unauditable, and — the part that stings in the review — still incomplete, because a human sampling the web is not a census of it.
When the client asks why a specific company is on the list, “the analyst thought it fit” is not an answer. A quoted sentence from the company's own site, with a URL, is.
Our position, stated as the client would want it stated: leading company databases index the companies they found;. we screen the entire active web —.
100M+ classified domains, 700+ industry categories, 99.99%+ of active internet usage —. against the mandate's acquisition criteria as written, then re-run on any subset when the criteria sharpen mid-mandate.
The mechanics run in two passes. Triage clears the category population down to live, in-scope operating companies — in one industrial specimen, a 25,000-domain US sample reduced to roughly 17,300 live operators.
Deep extraction then reads each survivor's site page by page and extracts fifteen defined signals, every claim carrying its verbatim quote and source URL.
The scoring lands where mandate reviews live: Mandate Fit is 70% of the score, computed against the criteria text itself; Outreach Suitability is 20%; Transition Context is capped at 10%.
Any company whose own site discloses group membership scores zero and moves to the exclusion log with the disclosure quoted.
Two findings from our specimen runs explain why this matters for an advisor's credibility: roughly one keyword-perfect candidate in ten was already group-owned, and a fifth or more of genuine fits never used the category's obvious keywords on their homepages.
The first statistic is the embarrassment your exclusion log prevents; the second is the “we found companies you didn't know” moment your renewal depends on.
The full methodology is documented openly on the method page.
The structure mirrors our public specimen format, scaled to the mandate. Four layers, each with a distinct job in the client conversation:
The ranked long list. Every confirmed fit, scored, with the evidence paragraph that lets a partner speak to any row in the review without notice. The hidden-fit layer. Companies matching the mandate whose sites never used the category's obvious vocabulary —.
flagged as such, because these rows are the visible proof of coverage beyond the client's own tools. The exclusion log. Keyword-plausible companies removed for documented cause: group ownership, wrong business model, out-of-scope geography —.
each with the disqualifying language quoted.
Reviews spend surprising time here, productively; nothing builds trust in the inclusions like honest exclusions. The insufficient-evidence flags. Companies whose sites would not support a confident verdict, labeled exactly that — not silently dropped, not guessed into the ranking.
In the public specimen this reads as 8 top fits, 5 keyword-missed fits, 5 documented exclusions, 2 flags; a mandate deliverable keeps the proportions honest at full scale.
Everything ships CRM-ready and unbranded by default — your mandate, your letterhead, your client relationship.
Scale context helps calibrate expectations.
In our specimen work, a 25,000-domain US-focused triage of one industrial category reduced to roughly 17,300 live operating companies, and single subverticals resolved to eligible-independent populations between 93 and 702 — numbers an advisor can quote in a review as the mandate's actual denominator.
The hidden-fit layer is typically a fifth or more of confirmed fits, which is worth stating to the client explicitly: it is the measurable share of their market that neither their tools nor their competitors' advisors are seeing.
And the analyst-verification pass is not ornamental — specimen rows carry notations like “15/15 evidence snippets verified against site text” because a partner should be able to read any row aloud in a client meeting without hedging.
All fifteen signals ship per company. In advisory work, these five get quoted in the room.
Scored against the mandate text, not a code table — which means the criteria your client actually wrote (“self-performing, multi-site, no pure distributors”) are what ranked the list.
When the client challenges a row, the answer is the criterion, the quote, and the URL, in that order. Reviews get shorter and better at the same time.
In regulated mandates the certification is the mandate: ASME stamps in boiler work, ISO/IEC 17025 in labs, AS9100 and NADCAP in aerospace supply, UL 508A in panel building.
We capture certifications as exact claim text, never as inferred attributes — so the compliance column your client's diligence team will rebuild anyway arrives pre-built and sourced.
Geographic mandates fail on the difference between an HQ pin and a real footprint.
We extract stated branches, service areas, and owned-versus-partner locations from the sites themselves — a four-location Florida operator is a different candidate than a one-office firm with a wide claimed radius, and the list should say so.
Nothing punctures a review like a candidate whose latest news item is five years old and whose phone number rings dead.
Activity trajectory — dated content, live careers pages, recent project posts — flags dormancy before the client meeting rather than after the first approach.
Mandates written as “service businesses” die in lists full of resellers.
The service-led/product-led/hybrid classification comes from what each site actually describes — field crews, contracts, install bases versus catalogs and carts — and it is the single most common reason a keyword-plausible company lands in our exclusion log for business-model cause.
Take a realistic mandate: a strategic acquirer wants independent boiler, burner and steam-system service companies, US, service-led, with recognized code credentials.
Our census of that category found 172 eligible independent US companies — a number worth pausing on, because it is small enough that completeness is checkable and every miss is visible.
The top of the ranking shows what evidence discipline produces.
A Florida single-source boiler house, four locations, carrying the ASME Power Piping National Board Repair “R” Stamp — fifteen of fifteen evidence snippets verified against site text.
A Tennessee operator across five locations, ASME-certified, whose about page traces the founder forming the company and a successor taking the president role. A Maryland family firm whose homepage states plainly:
And the exclusion log carries the row that wins the review: a keyword-perfect boiler rental and service company excluded because its own site announces it “was acquired by a global compressor manufacturer in June 2019.”.
An analyst-built list assembled from search results would have put that company in front of your client;. the census put it in the exclusion log with the quote attached.
That single row, shown in a mandate review, explains the method better than any slide about it.
Timing first: a full evidenced universe lands in two to three weeks, which advisory teams typically slot between mandate signing and the first client review — the window the analyst long-list build used to consume entirely.
The working pattern that has emerged among advisory clients: the deliverable arrives CRM-ready; associates enrich the top tier with contact data and the partner's market knowledge; the review presents the ranked list, the hidden-fit layer, and the exclusion log as three exhibits.
Client feedback in that meeting — a criterion tightened, a segment deprioritized — goes back into a custom-ICP re-run, included in the engagement, and the re-ranked universe returns within days rather than another analyst-month.
Longer mandates add quarterly re-screens so ownership changes and pivots surface before the client reads them in the trade press.
And the same machinery serves the other side of the desk: for sell-side work, buyer list construction inverts the question, screening for strategics whose sites evidence capability adjacency.
Firms running several mandates a year treat the whole thing as capacity — census work moves to the pipeline, judgment work stays with the team, and the associates who used to google candidate number 314 at midnight do something billable instead.
The highest-leverage hour of the engagement happens before any screening runs: converting the mandate as the client wrote it into criteria a pipeline can enforce.
Client mandate language is diplomatic by nature — “leading independent providers of critical industrial services in attractive regional markets” — and every adjective hides a decision. Does “independent” exclude ESOPs?
Is a company with one branch across the state line “regional”? Does “critical services” include distribution with a service desk, or only self-performing operators?
The intake works through each term with the advisor — not the client, unless you want us there — and writes the operational version down: the inclusion tests, the disqualifiers, and the deliberate ambiguities left open for the boundary band to illuminate.
This written thesis becomes three things at once.
It is the screening instrument — the text the model scores every site against, which is why criteria that exist in no database field still filter here.
It is the audit trail — when the client challenges an inclusion, the row cites the criterion and the quote, and the criterion cites the intake document the client saw.
And it is the change log: mid-mandate refinements are edits to a document, re-run against the standing universe within days, rather than a renegotiation of scope.
Advisors who have run several engagements report that the intake discipline improves their own mandate letters — once you have watched adjectives fail to filter, you write fewer of them.
A growing share of buy-side mandates are really compliance searches wearing M&A clothing: the client needs capacity that carries a specific accreditation, and the acquisition is the fastest way to own it.
These mandates reward the census approach disproportionately, because certification data in databases is thin, stale, and unsourced, while on company websites it is proudly specific — scope statements, certificate numbers, accreditation bodies.
Our specimen rows show the texture: an ASME Code Shop holding “the ‘U’. Stamp for fabrication, the ‘R’. Stamp for repairs, and the ‘PP’. Stamp for power piping”;.
labs accredited to ISO/IEC 17025:2017 with the accrediting body named;. NADCAP and AS9100D carried together on aerospace-facing machine shops;. UL 508A panel builders;. EASA-accredited motor rebuilders.
For these mandates the universe splits into evidenced holders, explicit non-holders, and silent sites — and the silent class matters, because certification-holding companies occasionally under-publish, and the mandate review should know the difference between “does not hold” and “does not say”.
The deliverable preserves exact claim text so the client's quality team can validate scope directly — an ISO 9001 shop and an ISO/IEC 17025 lab are different acquisitions, and keyword search conflates them constantly.
Several advisory clients now run certification-led mandates as a distinct product of their own, with the census as the engine and their regulatory judgment as the value layer.
Boundaries that protect the advisor as much as us.
No revenue or EBITDA estimates — websites do not contain financials, and a made-up number in a client deliverable is a mandate-level risk; sizing proxies we do extract (facility square footage, stated headcount, location counts) are labeled as exactly what they are.
No claims about owners' intentions or circumstances — our standards exclude that category of inference entirely, which matters because your client's counsel will eventually ask where the data came from, and every answer in our deliverable is “the company's own website, quoted.” No consumer-captive healthcare verticals.
And no pretense of covering companies with no meaningful web presence: in most B2B mandates that residue is small, and we report it as a stated boundary rather than extrapolating across it.
One email, the mandate criteria in a paragraph — the same-day specimen shows the evidence discipline your next review could run on.
Request the specimen report