What Makes a Company Worth Buying: The Criteria Behind a Real Screen

Every buyer has criteria. Most of them are three lines long — a revenue range, a geography, and a sector — and they exist mainly to be recited on introductory calls. A real screen is longer, more specific, and does something the three-line version cannot: it tells you what to reject, quickly, without a conversation.

The 2026 edition of Stanford Graduate School of Business's A Primer on Search Funds contains the most complete published version of such a screen for lower-middle-market acquisitions. It is written for search fund entrepreneurs, which means it assumes a first-time operator taking over as chief executive with capital partners behind them. Read the thresholds with that in mind; the logic beneath them generalizes further than the numbers do.

The size band, and the reason it has two edges

The Primer's stated financial criteria for a desirable company: $10 million to $30 million in revenue and greater than $1.5 million in EBITDA (p. 24). Companies below $10 million in revenue or $1.5 million in EBITDA are listed as undesirable — filed under "small company."

The lower edge is the one buyers expect. The Primer gives two reasons for it. The first is access to capital: "most lenders and investors favor" companies with at least $10 million in annual sales and $1.5 million in EBITDA (p. 25). The second is arithmetic. Below roughly $1 million in EBITDA, or with an EBITDA margin under 10%, the Primer calls a company dangerously small (p. 27), because there is no room to absorb a bad quarter, a lost customer, or the ordinary disruption of an ownership change.

There is a memorable version of the same point, attributed to a former searcher: "Even big growth on a small number still results in a small number" (p. 27).

The upper edge surprises people. Above roughly $30 million in sales and perhaps $7 million in EBITDA, the Primer treats a business as too complex for an inexperienced manager (p. 27) — multiple layers of management, more sophisticated systems, and a competitive set that includes better-resourced acquirers. The band marks the range where an operator can genuinely run the company they bought.

Within it, the Primer identifies a narrower zone: EBITDA of $1.5 million to $5 million is where "search fund entrepreneurs have had the most success" (p. 26).

Margin as a margin of safety

The Primer wants EBITDA margins greater than 15% (p. 27), describing that level as providing "a reasonable margin of safety," and flags margins under 10% as part of the dangerously-small profile.

The reasoning is about survivability rather than attractiveness. A 20% margin business that loses a fifth of its revenue is a smaller profitable business. An 8% margin business that loses a fifth of its revenue is a problem requiring immediate action from someone who has owned it for four months.

A related criterion sits alongside it: a return on tangible capital above 20%, which the Primer associates with "healthy and sustainable profit margins" at the industry level (p. 24). That one screens the industry, not the company — it asks whether the sector as a whole earns a real return on the assets it employs, or whether it is a business that consumes capital to stand still.

Concentration is the fastest disqualifier

The Primer's threshold: any single customer representing more than 30% of sales makes a company undesirable (p. 25).

This is the most decisive line in the entire screen, because it is binary, it is checkable relatively early, and it survives every other good quality a business has. A company with excellent margins, a durable niche, and a motivated seller is still a company where one phone call can remove a third of the revenue.

It is also the criterion most often argued away in month five, when the buyer is invested and the relationship with the customer is described as "thirty years, they're basically family." The Primer's broader advice about diligence applies: one serial investor described the discipline as needing to "separate the ants from the elephants," with the elephants being the issues that would cause a searcher to kill the deal (p. 39).

What counts as recurring

The Primer supplies a definition rather than a threshold, and it is stricter than most marketing materials assume. For revenue to count as recurring, customers must stay at least 18 months (p. 24).

This matters because "recurring revenue" has become a claim rather than a measurement in lower-middle-market materials. Monthly billing is not recurring revenue if the average customer leaves in seven months; that is a subscription-shaped way of describing churn. The 18-month test converts a marketing adjective into something you can check against a customer list.

Time horizon: the criterion nobody writes down

The Primer lists realistic liquidity options in three to six years as a desirable criterion (p. 24, restated p. 28).

Buyers screen on what a company is. Very few screen on whether anyone will want it later, which is a strange omission given that the exit is where the return is realized. The question is concrete: in five years, who plausibly buys this business, and why? If the honest answer is "another individual buyer, if credit conditions cooperate," the risk profile is different from a business that consolidators are actively acquiring in your region.

Industry before company

The Primer devotes an entire section to evaluating industries before evaluating businesses, and describes a funnel for it. One former searcher advocates starting from a top-down list of approximately 75 industries (p. 29). Initial screening — a process the Primer says takes one to two months — narrows that to five to 10 promising industries (p. 31). Information gathering narrows it further to a top three.

The Primer also records a simpler approach some searchers take, screening on only two or three "super-priority" criteria — for instance considering only industries with recurring revenue, the ability to scale, and at least 20 potential targets, with all other industries immediately eliminated (p. 31). An industry that cannot clear that count is not searchable, however attractive the businesses in it are, because a search that depends on two specific companies saying yes is not a search.

There is a time-management warning attached: "Spending three days each on 10 industries can easily result in a month of effort" (p. 32). Industry work feels productive and consumes the same months that sourcing needs.

Sectors that punish good operators

The Primer's position on declining sectors is unambiguous:

Historically, search funds have not done well "being the best house in a bad neighborhood"; declining industries often deteriorate into a zero-sum game, with increasing pressure as companies compete primarily on price.

— Primer, p. 26

It reinforces the point with Warren Buffett's line — "When a management team with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact" — and its own conclusion, which is the sentence most worth arguing with before you disagree with it:

Experience in the search fund industry has shown that principals have been better off paying full price for a good company than getting a "bargain" for a bad one.

— Primer, p. 27

The Primer also notes, without prescribing a threshold, that technology exposure is now a live screening question. Chief executives at a recent conference rated the urgency of artificial intelligence to their businesses across a range from 10 out of 10 to under 5 out of 10, but all of them were investing in AI literacy (p. 28).

Criteria are a rejection tool, not a wish list

The failure mode of a long screen is that it becomes a description of a company that does not exist. The Primer addresses that directly:

A key challenge facing search entrepreneurs is to know "when to take a train" and acquire a business, lest they never leave the station by waiting for opportunities that perfectly fit all of their criteria.

— Primer, p. 23

The resolution is to sort your criteria into two groups before you start, not while you are staring at a specific company.

Disqualifiers are conditions where the answer ends the conversation regardless of everything else: customer concentration above 30%, a structurally declining industry, an EBITDA margin in the single digits, a business that cannot be run by the person buying it. These should be applied early, mechanically, and without sentiment.

Preferences are everything else — geography, growth rate, the specific end market, whether the owner will stay through a transition. These trade off against each other and against price, and they are what "taking a train" means.

Most buyers hold their criteria the other way around: rigid on preferences, negotiable on disqualifiers. It is the exact inversion, and it produces the outcome the Primer describes as the worst available: buying an unattractive business "that limps along without the opportunity for a rewarding or graceful exit" (p. 10).

The hierarchy worth memorizing

An experienced searcher and investor gave the Primer the cleanest statement of what a screen is for:

A searcher's hierarchy of outcomes is (1) buy a good business, (2) don't buy a company, and (3) buy a bad company.

— Primer, p. 28

Ranking "no acquisition" above "wrong acquisition" is what makes a screen worth having. If the second and third outcomes were reversed, criteria would just be preferences and every one of them would eventually bend.

Turning this into something you can run

A usable screen has three layers, applied in cost order.

The first layer is checkable without contacting anyone: sector, size band, geography, and whether the business is structurally the kind you could run. Most companies exit here, and they should — this is the cheapest step and it should carry the most volume.

The second layer requires evidence but not cooperation: customer concentration signals, industry trajectory, ownership tenure, whether a succession event is plausible. This is where a real screen distinguishes itself from a three-line one, and it is the layer most buyers skip because it takes work per company.

The third layer requires the owner: verified financials, actual customer concentration, the reason for selling, and price expectations. It is expensive, and it should only ever be reached by companies that cleared the first two.

The Primer's screen is worth adopting as a starting point and worth adjusting to your own capital and capability. What is not worth adjusting is the structure — a small number of hard disqualifiers applied early, a larger set of preferences applied late, and a written commitment to the idea that not buying is the second-best outcome rather than the worst one.

See how DealPort turns acquisition criteria into a repeatable, evidence-backed screen.