Call him Jim. He’s 63, he has a CPA practice doing about $1 million a year, and he wants to retire. He built the book himself over two decades. His clients like him. His margins are normal for the profession. By every measure a small-business owner would recognize, he owns something valuable.
He also can’t sell it — not through the channels the trade press writes about, anyway. Jim is a composite of a lot of conversations, but the arithmetic behind his problem is exact, and almost nobody explains it to him before he starts.
The arithmetic nobody explains to the seller
There are roughly 87,000 accounting firms in the US, and most of them are one CPA and some help. Somewhere between 30,000 and 40,000 of them, carrying more than $20 billion of annual revenue, are expected to change hands as their owners retire. AICPA data put about 75% of CPAs within 15 years of retirement. The Rosenberg survey has 65.5% of partners at $2–10 million firms over the age of 50, and 71.1% among sole practitioners. This is not a wave that is coming. It is the current state of the profession’s ownership.
Now put Jim’s firm against how capital actually prices these deals. At a normalized 30–35% margin, $1 million of revenue is $300,000 to $350,000 of EBITDA. Lower-middle-market private equity generally starts underwriting at $2 million of EBITDA. Sponsors building a platform commonly want $5 million before they’ll treat a firm as the base of one. Pepperdine’s 2025 Private Capital Markets Report documents exactly this shape: capital is short below $5 million of EBITDA and in surplus above $10 million. Senior debt gets materially harder to arrange under $10 million.
Jim is at a third of the lowest of those numbers. And he is not the exception — 81% of the responses in the AICPA’s MAP survey came from firms at $5 million of revenue or less. Jim is the profession’s baseline, not its tail.
Three frictions, and only one of them is about size
The asset doesn’t convey at closing. On a spreadsheet, Jim’s revenue looks investment grade: recurring, low attrition, predictable. What the spreadsheet doesn’t show is that the recurrence runs through one person. A buyer isn’t underwriting the revenue, they’re underwriting the probability that it survives Jim’s departure. Low historical attrition is not evidence about that question. It is evidence that Jim never left.
The successor pipeline that used to absorb these firms is gone. The traditional exit was a younger accountant buying a book to build their own practice on. More than 300,000 accountants and auditors left the profession between 2020 and 2022, and enrollment has been under pressure since. Fewer younger practitioners means fewer of exactly the buyer Jim was counting on, and the ones who remain have their own capital ceiling.
Diligence costs the same whatever the target earns. A quality of earnings report runs $15,000 to $25,000. Legal review adds $5,000 to $15,000. The IBBA and M&A Source Market Pulse reports put diligence at three to four months after a letter of intent. That spend is essentially fixed, and around 31% of sale processes never close at all. So a buyer stakes the same money and the same four months against $325,000 of EBITDA as against ten times that. Where practices run on undocumented manual work — document collection by email, categorization by memory, reconciliation in a spreadsheet, workpapers in a folder only the owner understands — that spend also buys less certainty, because there is no clean earnings history to verify.
The real problem is a missing bidder
Here is the part I find most worth saying, because it cuts against how sellers usually read the situation. Jim’s firm is not worthless. It is unpriced. Nobody has run a real competition for it.
A consolidator would very likely pay more for Jim’s book than a local CPA would — something like $1.5 million where the local buyer tops out near $1.2 million for the identical firm. Not because consolidators are generous, but because they’re buying a component. They fold the back office into infrastructure that already exists, cross-sell advisory into the client base, and underwrite the multiple against a bigger machine. A local CPA is buying a job. Their ceiling is set by personal income capacity and what a bank will lend against it, which in practice caps them near three years of earnings no matter how much they want the firm.
What actually happens instead: practices at Jim’s size draw about 2.2 offers per deal, against 3.9 above the threshold, and roughly 60% of buyers sit within 20 miles of the seller’s office. Two bidders from the same county is not a market. It is a negotiation the seller has already lost, and it shows up in every term that matters — price, transition length, retention guarantees, what happens if clients leave.
Four things that actually widen the market
Merge before you go to market. Two or three practices combined ahead of a sale can clear $1 million of EBITDA together. That is a different conversation entirely: the group lands on the radar of add-on buyers who would not have opened the file on any of the firms individually.
Go to platforms that are already funded, not to sponsors looking for one. A sponsor hunting a platform won’t buy Jim’s firm, because it doesn’t meet the minimum. A platform that sponsor funded two years ago explicitly wants tuck-ins Jim’s size, and underwrites them on completely different math, because the back office and the technology are already paid for.
Run the process wider than your zip code. The local concentration of buyers is not a law of nature. It is what happens when an owner sells the way owners have always sold, through local intermediaries and word of mouth. Deal platforms that aggregate lower-middle-market buyers get used routinely above $5 million of enterprise value and are barely used below it. Above the threshold, more than half of buyers are more than 100 miles from the seller.
Modernize the operations, fast. Going fully remote with clients, getting paper out of the workflow, and standardizing on a current technology stack does two things at once: it cuts the integration friction a buyer has to price in, and it opens the firm to a younger, tech-comfortable buyer who wants a clean book without inheriting somebody’s legacy infrastructure.
Why the timing is the whole thing
Every one of those moves depends on the practice running institutionally rather than personally — documented onboarding, standardized workpapers, one technology stack everyone actually uses. That work takes time, which is the argument for starting it now rather than at the point of sale.
The instinct is usually the opposite: wait a couple of years, let the retirement wave build, sell into a hotter market. It doesn’t work that way. The wave lengthens the queue of sellers, it doesn’t clear it. And the local buyers who were available to Jim this year will have committed elsewhere by the time he’s ready.
If Jim’s arithmetic looks like yours, the useful next step is not a valuation. It’s deciding which of those four moves your firm is closest to being able to make, and what has to be true operationally before it can.