Writing · The market

Why 92% close instead of selling.

McKinsey counted about 510,000 US small-business exits in 2022. 92% were closures. 5% were sales. 3% were transfers. Here is what is behind that, and which reasons are fixable.

Read the numbers again, because they are worse than they look.

Out of roughly 510,000 American small businesses that came to an end in 2022, about 92% simply closed. The doors shut. The equipment went to auction, or into a barn. The phone number went dead. Around 5% were sold. Around 3% were handed to family. That is McKinsey, published in February 2026.

Most owners do not fail to sell. Most owners never get to the question.

I want to be straight about one thing before I go on. For years the standard line in this business was that most listed companies never sell. I used to use it. It does not survive checking — it traces back to one quote of one small survey, and the same group publishes very different figures elsewhere. So I retired it. The McKinsey number is better sourced and it is a bigger problem, which is a strange thing to be pleased about.

Why closing is the normal ending

Because most of these companies were never assets. They were jobs with equipment attached.

That sounds harsh. It is not meant to. A man builds a plumbing company over 30 years. He knows every customer, prices every job, holds every relationship, and the whole thing works because he is in it every day. He has built something real and he has fed a family with it. But there is nothing there that can be handed to another person, because the thing that makes it work is him.

What cannot be handed over can only be switched off. That is the 92%.

Of the ones that do try

Now take the small share who get as far as listing. The odds are still poor, and they get better as the business gets bigger.

And the reasons, in order of how often they are named: the price the owner expects, at about 35%. Financial records that do not hold up, at about 25%. A business that leans too hard on the owner, at about 20%.

Every one of those three is fixable before you list. None of them is fixable after.

Which reasons are fixable, and how long each takes

The price you expect — fixable, and the cheapest fix there is

Most owners have a number in their head. It came from a competitor's sale ten years ago, or a magazine, or what the retirement needs to be. Nobody has ever tested it.

Getting a real reading two years early costs very little and changes everything downstream. It either tells you the number is close, which is a relief, or it tells you there is work to do while there is still time to do it. I do not give valuations of record — that is an appraiser's job, and there is a point where you should hire one. But knowing what a buyer's bank will actually see is not a valuation. It is arithmetic, and you can have it years early.

Books that do not hold up — 12 to 24 months

There are two kinds of books: the kind that satisfies the tax man, and the kind that survives a stranger's accountant. Most businesses between $2M and $5M have the first kind. Nothing wrong with that, until money is on the table and every soft spot turns into a question.

And here is the part owners get wrong. Weak books do not knock a percentage off your price. They change who is allowed to buy you. With books a bank can read, your buyer is anybody with a bank behind him. Without them, your buyer is somebody paying cash. At a price between $900,000 and $2.5M, that is not a discount. That is the loss of the market.

A business that leans on the owner — 12 to 18 months

Same shape of problem. "There is an operator besides the owner" is a box on a credit memo, and an empty box is a reason to decline. Getting your owner-touch count down from 20 a week to under five is a year of deliberate handing over, one decision at a time.

One customer too big — 24 to 36 months

The slowest of the four, which is why it has to start first. Above roughly 15% to 25% of revenue in one customer, whole categories of buyer stop looking. Above 30%, most stop entirely. It is a gate, not a discount, and the only ways through it are to grow everything else or to move some of that work somewhere else. Both take years.

And the ones that are not fixable

I will name these plainly, because pretending otherwise is how consultants lose people.

• Health that arrives early. A stroke in the spring does not care about your 18-month plan. This is the whole argument for starting in a normal year.
• A trade the buyers have left. Some lines of work have no pool of buyers at any price, and no amount of tidy books creates one.
• A location you do not control. If the lease is short and the landlord will not extend, the buyer is buying a moving job.
• A partner who will not sell. That is a legal problem and a family problem, and it needs an attorney, not me.

And then there are the deals that die after the handshake

This is the part that surprises owners most. Agreeing on a price is not the finish line. It is the start of the examination.

Axial looked at 75 letters of intent that fell apart in 2025. Roughly 47% of them died inside the buyer's review — after both sides had already agreed on a number. Of the deals that break after that handshake, about 25.3% break on what the review turns up, and about 21.3% break on earnings that do not match the story. That second figure has doubled: it was 10.6% in 2023.

Which tells you something useful. Books that would have survived a buyer's review three years ago do not survive one now. The bar moved.

What the 92% actually costs a family

Not a lecture. Just the arithmetic of a closure.

Equipment sells at auction for what auctions pay. The trucks go individually. The customer list is worth nothing, because a list with no company behind it is a spreadsheet. The name on the sign is worth nothing the day after the sign comes down. Whatever the business was worth as a working company, none of that ever becomes money.

Meanwhile the crews find other jobs, which is fine, and the customers find other suppliers, which is also fine. The only person who loses the whole thing is the owner.

The clock is the real problem

Look back at the timelines. Books, 12 to 24 months. Owner dependence, 12 to 18. Concentration, 24 to 36. Price expectation, whenever you are willing to hear it.

Every single reason on the list takes one to three years to fix. Almost nobody starts the clock on time, and the ones who never start it at all are the 92%.

The good news, and it is genuinely good: all of this work pays you while you do it. Better pricing, faster cash, fewer callbacks, a second-in-command who covers your Saturdays. If you never sell a thing, you still built a better company and got your weekends back. That is the only readiness work I know of where the worst case is still a win.