What a small business actually sells for

There is no single answer to what a small business is worth, and the reason is not that valuation is subjective. It is that "small business" covers markets so different from each other that the same word describes a shop selling for 350,000 dollars at under three times its cash flow and a company selling for 16 million at over six times its earnings. Whoever quotes you a multiple is quoting from whichever of those markets they work in.

Three datasets, three different worlds

Rather than argue about it, put the actual numbers side by side. Each of these is a real transaction record rather than an opinion, and each covers a genuine segment of what people call small business acquisition.

BizBuySell tracks completed sales of American main-street businesses. Across 2025 it recorded 9,586 closed transactions, with a median sale price of 350,000 dollars, median revenue of 703,000 and median cash flow of 158,950. By the second quarter of 2026 the median price was 349,250, at a cash flow multiple of 2.7 and a revenue multiple of 0.7.

Stanford's search fund study covers a different animal entirely. For the 2024 to 2025 cohort the median acquisition had an enterprise value of 16 million dollars, on median earnings of 2.5 million, at a multiple of 6.2 times.

And in the German-speaking market, the DUB multiples, aggregated quarterly from more than twenty-five M&A advisors and financial institutions, put Mittelstand EBITDA multiples across a range of roughly 2.4 to 10.9 times depending on sector and size. Software and comparable businesses sit at the top, around 7.7 to 9.7. Trades and consumer goods sit at the bottom, around 2.4 to 5.5. Typical micro-cap deals run four to seven times, mid-cap five to ten.

One caution on reading that table, because it is the mistake it invites: medians do not multiply. BizBuySell's median price is 350,000 while its median cash flow times its median multiple comes to about 429,000, and neither figure is wrong. They are separate medians of separate distributions, so use each column as a description of its own market and never as a calculator.

The size effect, which nobody explains to owners

Read those three together and the pattern is unmistakable. As earnings rise, the multiple rises too, which means value scales faster than profit. Double the earnings and you have more than doubled the price, because you have also moved up a multiple band.

The size of that effect is worth stating exactly rather than as a tendency. Between the two ends of the table, earnings are roughly 15.7 times larger and the multiple is 2.3 times higher. Those compound, so the price is not 15.7 times bigger. It is around 36 times bigger. Somewhere between a third and a half of that gap has nothing to do with the profit at all. It is paid for being the size where a different set of buyers turns up.

The reasons for the band shift are structural rather than sentimental, and they are worth knowing because each one is something you can deliberately move toward.

1. The buyer pool changes

   Below a certain size the only buyers are individuals purchasing a job for themselves, and they can only pay what a bank will lend against their own covenant. Above it, funded buyers and other companies enter, and they compete against each other.

2. Management stops being the owner

   A company with a layer between the owner and the work can be bought by somebody who does not intend to run it. That single fact roughly doubles the number of people who can bid.

3. The numbers become verifiable

   Larger companies have audited or reviewed accounts, real systems and a finance function. Verifiable earnings are worth more than the same earnings requiring a leap of faith, and the discount for unverifiable is severe.

4. Fixed deal costs stop dominating

   Legal, diligence and advisory costs are broadly similar on a 400,000 deal and a 4 million one. On the smaller deal they are a punishing percentage, and that shows up in the price a buyer can justify.

Getting a management layer in place has the best return on effort. Getting yourself out of the daily operation is usually described as a lifestyle improvement. It is also the single change most likely to move a business from one multiple band to the next, because it changes who is allowed to bid. What that actually takes is in owner dependence.

Sector matters, and less than people think

The DACH spread from 2.4 to 10.9 looks like sector is decisive, and it does matter: a software business will always price above a trade business, because the earnings are more scalable and less tied to people who have to physically arrive somewhere.

But comparing across sectors is mostly a distraction for an owner, because you cannot change what industry you are in. Within a sector the spread is still wide, and everything driving that spread is something you control: how much of the revenue is contracted, whether the customer base is concentrated, whether the business runs without you, and whether the accounts can be verified.

You cannot become a software company. You can become the version of your trade business that prices at the top of its band instead of the bottom, and the distance between those two is larger than the gap between sectors.

There is a second reason not to fixate on the sector average. These published multiples are ranges assembled from many transactions, and the individual deal you do will land somewhere inside the range based on the specifics of your company. A published figure tells you which band you are negotiating in. It does not tell you your price.

How to read a multiple somebody quotes you

Five questions will tell you whether the number you have been given means anything.

• What is it a multiple of: revenue, EBITDA, or an owner cash flow figure that already includes the owner's salary
• Which market does the source operate in, because a main-street broker and a mid-market advisor will quote very different numbers and both will be right
• Is it drawn from completed transactions or from asking prices, since the gap between the two is real and always in the same direction
• Does it include the property, the vehicles and the stock, or is it for the operating business only
• Is working capital inside the number or provided separately, because that alone can move the real cost by a large margin

The first is where most confusion originates. Main-street data typically uses a cash flow measure that adds the owner's salary back, so a 2.7 multiple of that is not remotely comparable to 6.2 times an EBITDA figure that has a market salary deducted from it. Those two numbers describe similar businesses and are not on the same scale. Rebuilding earnings so they are comparable is its own exercise, and it is the one that has to happen before any multiple means anything.

What this changes for a buyer

Two things, and the second is the one that shapes a strategy rather than a deal.

First, know which band you are buying in and do not import a multiple from another one. Paying six times for a 200,000-earnings business because a mid-market report said six times is the going rate is a straightforward way to overpay by a factor of two, and it happens constantly.

Second, and more interesting: the size effect is the entire arithmetic behind consolidation. Buying several businesses in the four-to-seven band and running them as one company with combined earnings in a higher band creates value from nothing but scale, before any operational improvement at all. That is real and it is why the strategy exists. It is also why so many attempts fail, because the combined entity has to actually function as one business rather than as a holding structure with several separate companies inside it, and that is the hard part.

For what it is worth, I think the multiple bands are the least interesting thing in this whole subject and the most discussed. The number that decides whether an acquisition works is not what you paid relative to a published range, it is whether the earnings were real and whether they survive you owning the place. A fair price for a durable business beats a bargain on a fragile one every time, and the published tables cannot tell you which is which. What I actually check first is on the buy page.

US transaction data from the BizBuySell Insight Report (2025 full year and Q2 2026). Search fund figures from the 2026 Stanford Search Fund Study. DACH ranges from the DUB KMU Multiples, aggregated quarterly from over 25 M&A advisors and financial institutions.