← All articles

Business Succession: The Know-How in the Owner's Head

When a small business changes hands, the know-how that lives only in the owner's head costs the seller money and gives the buyer a risky first year. AI makes it cheaper to write it down in time.

An open leather notebook with a ring of brass keys on a worn workshop desk, standing in for the know-how an owner hands over in a business succession

Key takeaways

  • Germany expects around 186,000 company handovers between 2026 and 2030; France has about 370,000 businesses potentially up for sale within five years, but only about 26,000 changed hands in 2024.
  • Buyers pay for what keeps working after the owner leaves. Knowledge that lives only in the owner's head shows up as risk: a lower price, a longer earn-out or a longer forced stay for the seller.
  • AI makes it much cheaper to get that knowledge on paper before the sale: draft the owner's rules from existing emails and quotes, fill the gaps in conversation, and test the result the way a buyer would.
  • For buyers, the first 100 days are for mapping how the business really runs. Automate the repetitive work around the experts, and keep the people who hold the customer relationships.

Ask the owner of a 25-person company how a quote gets priced and you often get a shrug and "I just know." That answer is fine for 30 years. In a business succession it becomes a line item: the buyer prices in the risk that the knowledge leaves with the seller.

Why more businesses are changing hands now

The numbers are large on both sides of the Rhine.

In Germany, IfM Bonn counts around 186,000 companies due to be handed over between 2026 and 2030. KfW's succession monitor puts planned successions at about 109,000 a year until the end of 2029, and more than one in two owners is 55 or older.

In France, a 2025 study by Bpifrance Le Lab with CCI France, CMA France and the CRA surveyed about 5,000 business leaders. It found about 370,000 businesses potentially up for transfer within five years. In 2024, about 26,000 actually changed hands.

That gap is where a lot of value gets lost. Some businesses close because nobody takes them over. Others sell, but for less than they're worth, because the buyer can't see how the business will run without the person who built it.

The buyers are changing too. Next to family successors and managers buying in, there are search funds, small private equity firms and, more recently, "AI roll-ups": investors who buy several businesses in one field and rebuild how they operate with AI. Buying and combining small firms is an old pattern. In the US accounting market alone, the CPA Trendlines deal tracker counts 177 initial private equity investments between 2015 and 2025, followed by 875 add-on acquisitions. What AI changes is the size of business that becomes interesting to buy and how much of the back office a buyer expects to rework.

For an owner thinking about selling, all of this points to the same question: how much of my business works without me?

What the owner knows that nobody wrote down

Take an illustrative example: an electrical contractor with 25 people, run by the founder for 28 years. The accounts are clean. The order book is full. The value sits somewhere less visible.

  • Pricing: the founder knows which jobs to quote high because the building is old and the wiring always hides surprises. That knowledge is in no spreadsheet.
  • Customers: one property manager wants a phone call before every invoice. Another only pays if the PO number is on page one. A third is worth keeping despite late payments because they send the big refurbishment jobs.
  • Suppliers: which wholesaler can deliver the same day if you call before 10. Who gives a better price at quarter end.
  • People: which technician the difficult customers trust. Who should never be sent to the school jobs again.

None of this shows up in due diligence documents. A careful buyer still senses it, usually in the first meeting, when every answer starts with "I usually handle that myself."

How buyers price that risk

Buyers rarely say "your knowledge isn't documented, so we'll pay less." It shows up in the deal terms instead:

  • A lower multiple, because future earnings look less certain.
  • A bigger share of the price paid later as an earn-out, tied to results after the handover.
  • A longer period in which the seller has to stay on and keep the business running.

For a seller who wants to retire, that last one hurts most. The more the business depends on them, the longer they can't leave.

We wrote about the same problem from a different angle in software succession planning for small AI vendors: when one person holds the knowledge, the customer carries the risk. In a business sale, that risk gets a price.

Getting owner knowledge on paper before the sale

Documenting a business used to mean weeks of interviews and a binder nobody opened again. That's why most owners never did it. AI makes it much cheaper, and for a seller the goal is narrow: show a buyer that the business runs without you.

Start from what already exists

Years of emails, quotes, invoices and job notes already contain most of the patterns: which customers get which terms, how quotes are built, which jobs ran over and why. AI can read through them and draft the rules the owner follows without thinking. The owner then corrects the draft, which is far quicker than writing from a blank page.

What was never written down comes out best in conversation. The owner talks through a quote, a difficult customer or a supplier deal, and the recording becomes a short written procedure. Ten sessions like that often cover more than a month of trying to write a handbook.

Test it the way a buyer would

Give the result to someone who doesn't know the business, with ten real questions from last month. Where they get stuck, a gap remains. Do this while the owner can still fill it: six months before a sale is a good time, six weeks after is too late.

Keep it in the company's name

Whatever gets captured should sit in systems the company owns and hands over with the shares, so it can't disappear with the owner's personal accounts or a lapsed subscription. As we argued in you don't need to build your own AI, what matters is owning what the AI learns. In a sale, that ownership is part of what the buyer pays for.

What a buyer should automate in the first 100 days

The other side of the deal has its own temptation. AI makes it easy to imagine cutting half the back office in the first quarter. In small businesses, that tends to break exactly the relationships the buyer paid for.

A process-first approach runs in three steps.

Days 1 to 30: map before changing anything

Follow how a job really moves through the business, from first call to paid invoice. Where does it wait? Who does it depend on? Which steps are repetitive, and which need judgement? The written procedures from the seller help, if they exist.

Days 30 to 60: take repetitive work off the experts

The best first targets are the tasks that steal time from the people customers value: sorting incoming requests, drafting standard quotes from the captured pricing rules, chasing overdue invoices, preparing job reports. The technicians and the account manager keep doing the work only they can do, with less admin around it.

Days 60 to 100: measure and decide

Set a baseline before changing anything: response time to customers, time from job to invoice, overdue receivables, hours spent on admin. Compare after. Keep what moved the numbers. Drop what didn't.

The people to keep close are the ones customers ask for by name. In a 25-person business, that's often three or four people. AI can make them more productive. It can't replace why customers trust them.

A note for roll-up investors

Roll-up investors bring something many small businesses lack: capital, shared systems and people who have seen the same problems in ten similar companies. Done well, that's good for the businesses they buy, and for owners who would otherwise close without a successor.

The roll-ups that will work best are likely the ones that treat each acquired business as a set of relationships and know-how first, and a cost base second. The knowledge capture above is useful to them too: it's what lets a good practice from one acquired company spread to the others.

The part of the business that isn't in the accounts

Every year, thousands of good small businesses in Germany and France either close or sell below their value because what made them work was never written down. AI makes it much cheaper to show a buyer that the business keeps working after the founder hands over the keys, and that is often what decides the price.

Frequently asked questions

Ideally one to three years before. Knowledge capture can be done in a few months, but you want time to test it, to show the business running with less of your involvement, and to fix gaps before a buyer finds them.

It can do most of the heavy lifting. AI reads existing emails, quotes and invoices for patterns, and turns recorded conversations with the owner into written procedures. The owner still has to check the result. What AI removes is the weeks of writing that stopped most owners from ever doing it.

Anything that touches the relationships customers value: the call before a big quote, the personal follow-up after a problem, the technician a customer asks for. Automate the admin around those moments and leave the moments to people.

There's no fixed formula, so be wary of anyone promising a percentage. What documented processes do is reduce the buyer's risk, and buyers price risk through the multiple, the earn-out and how long the seller must stay. Less dependence on the owner gives the seller a better position on all three.

No. The smaller the business, the more of it tends to sit with one person, so the effect of capturing that knowledge is usually larger.

Want to see what AI can do for you?

Tell us about your business. We'll get back to you within 24 hours.

Schedule a Strategy Call

More articles