Open any SERP for "what makes a good acquisition target" and you get the same nine bullets in a different order. Strong management. Recurring revenue. Strategic fit. Cultural compatibility. Clean cap structure. Diversified customers. Defensible margins. Growth runway. Reasonable price.
None of it is wrong, and almost none of it is something you can score before the second meeting.
From the vantage of someone who'd have to build the scrapers and scoring underneath this, the criteria reads like a diligence checklist someone retitled. Half the items only resolve once you're inside the data room. The other half resolve in a conversation. If a criterion can only be evaluated after you've already paid the cost of getting in front of the seller, it isn't a criterion. It's a wish.
The cost of treating wishes like criteria is real. A partner spends Tuesday calling ten companies that look great on a recurring-revenue proxy that doesn't hold up. An associate burns three weeks on a target whose owner sold ten years ago. The firm pays for a data subscription whose scoring can't tell those two apart. Multiply that across a year and you've underwritten a full headcount of motion that produced nothing.
The useful question is which subset of the standard list you can actually score on a Tuesday afternoon from public data, and what the score is allowed to predict.
The three upstream buckets
Pull the standard checklist apart by "can this be measured before the first call," and it collapses into three real buckets. Everything else is diligence.
Operational signature. What the company actually does, at what scale, how repeatedly. Headcount over time, role mix (sales-to-ops ratio, whether there's a controller, whether there's a real engineering function), how many sites they run, permit volume in regulated trades, fleet size, license counts, store count. A careful scrape will get you to a confidence interval that's useful. These aren't revenue, but inside a given vertical they track it closely enough to stand in as the revenue proxy when no third-party estimate holds up, which in the lower middle market is most of the time. The vendor estimates here are usually a model running on the same inputs anyway.
Financial proxies, not financials. You won't get audited financials before a conversation. You'll get the things that imply them. Years in business, registered fleet count and asset filings, commercial real-estate footprint, Better Business Bureau (BBB) and license-bond history, court records, Uniform Commercial Code (UCC) filings, whether there's a working capital facility on file. None is the income statement. Together they draw a sharper picture of a company's scale and durability than a vendor's revenue estimate does.
Ownership profile. Who owns it, for how long, roughly how old they are, whether they have a second business, whether they took outside capital. The public-company-flavored checklists ignore this bucket most aggressively, because public-company writeups assume a board and a fluid cap table. In an owner-led private business, the ownership profile is the criterion. A great financial performer with a forty-two-year-old owner who took growth equity last year is not the same target as the identical company with a sixty-three-year-old founder who never raised. One of those is a target you can buy in the next eighteen months. The other isn't.
That's the working scorecard. Operational signature, financial proxies, ownership profile. Three buckets, each of which you can populate from public sources, before anyone has picked up a phone.
The diligence-variable problem
Cultural fit, management strength, customer concentration, recurring revenue percentage, defensibility. All of these are real, and none of them are upstream-measurable.
Management strength is a diligence variable. You learn it in the room. The closest upstream proxy is tenure and pattern of hiring, already inside operational signature.
Customer concentration is a diligence variable. The closest upstream proxy is industry mix from press releases and case studies, which overfits to the companies whose marketing teams write better case studies.
Recurring revenue percentage is a diligence variable. The closest upstream proxy is the business model class (subscription vs. project vs. transactional), which you can usually infer. The percentage itself can't be scored upstream with any accuracy. Anyone claiming they have is showing you a guess that the firm is paying to dress up.
Strategic fit isn't a property of the target at all; it's a property of the buyer-target pair, and it belongs on the buyer's checklist rather than the universe-of-targets scorecard.
This is where the cost lands. A pipeline that tries to score on diligence variables either over-filters and rejects good targets because the proxies are noisy, or waves through bad targets that happen to look clean on those same noisy proxies. Both failure modes get paid for in partner hours, which is the most expensive thing a firm has.
How the three buckets trade off
A few rules of thumb if you're trying to weight them.
The buckets are not equal. For any firm buying owner-led businesses, ownership profile is the heaviest of the three, because it controls whether the company can be bought at all in the next eighteen months. Operational signature controls whether it's the right size and shape. Financial proxies are the tiebreaker. Score all three at equal weight and you'll end up with a list that's correctly sized and operationally fit, a meaningful chunk of which has no path to a transaction because the owner is forty and just raised.
A weak score in one bucket is not redeemable by a strong score in another. A great operational signature with a no-signal ownership profile is a target you'll spend a year on without ever finding a moment to buy. A great ownership profile with a weak operational signature is a tuck-in at best. The buckets multiply rather than add.
The vertical sets the weights. The signal groups that carry the most predictive load in home services are not the ones that carry it in ag services or specialty chemicals; generic weights across industries don't survive a backtest. Whichever vertical you're sourcing in, the weights have to be fit on data inside that vertical, not borrowed.
What this looks like in a pipeline
The pipeline is the artifact, not the list. Operational signature comes from registries, job-board histories, permit data, and site scrapes. Financial proxies come from corporate filings, UCC and court records, asset registries, and the BBB. Ownership profile comes from the same registries cross-walked against leadership pages and local press, with age estimation from public records rather than LinkedIn, which misses the population an owner-led-business thesis depends on.
Each input has a freshness window. Permits update on a roughly weekly cadence in most jurisdictions. Job-listing data updates roughly daily. Corporate filings update on a state-by-state cadence that has to be modeled separately for each Secretary of State. The score is only as good as the freshest input, which is why what's worth owning is the pipeline rather than the list.
A target list is not a deliverable; it's the output of a system, on the day the system ran. If the deliverable is the list, you bought a snapshot that's stale inside a month. If the deliverable is the system, you own the answer every Tuesday, against data that has moved. The criteria are the easy part. The system that scores ten thousand operators against them, this week and the week after, is the part worth paying for, and it's the part the standard checklist quietly assumes someone else is building.
