Recurring revenue has long been considered proof that a company does not depend on the next deal. Subscriptions, contracts, fixed maintenance clients: revenue that comes back without new sales effort. A buyer counts that revenue more heavily than one-off assignments, and that does not change.
What does change is the question behind it: recurring revenue delivered by what? If the answer is a team that manually produces the same reports, invoices, or checks every month, then the buyer is not only buying the revenue. He is also buying the labour costs needed to keep delivering that revenue. If the answer is a process that largely runs without human hands, he is buying something else: revenue with a lower cost per client, one that grows without headcount growing proportionally.
So the buyer is not buying the number of people behind the recurring revenue, but what keeps running without those people. That is the core of why this driver now weighs differently than it did five years ago.
Work that AI takes over does not only change costs. It changes the quality of the revenue itself. Three categories run through this:
1. Work that AI takes over completely: a report that is automatically compiled and sent without waiting on an employee. 2. Work that AI partly takes over, with a human approving or rejecting for a reason: a draft invoice that an employee checks before it goes out. 3. Work that remains human work: the conversation with a client who wants to change his contract.
A subscription portfolio built on category 1 and 2 has a different profit quality than a portfolio that relies entirely on category 3, even if the revenue is identical on paper. The buyer sees that difference reflected in the margin that remains after the acquisition, not in the revenue itself.
This is already happening today, and not evenly everywhere. An accounting firm with fixed clients can largely have the year-end close prepared by software and only review it itself; a consultancy with fixed retainers has the same repetition, but the work itself remains largely conversation and judgement. The difference is not in the sector as a whole, but in what the work within that recurring revenue actually consists of.
A buyer rarely asks directly about AI. He asks about things that reveal the answer:
That last point touches on the question of how a company would function without the owner or specific employees, and that question is explored further on the page about why a buyer looks at the management team. A similar logic applies to recurring revenue: the more that revenue depends on a system rather than on a person, the higher a buyer values it.
There is no percentage that applies to every company; that depends on the sector, the type of contract, and how the work behind those contracts is organised. What can be measured is the following:
This measurement inevitably touches on the question of which work in your company can actually be taken over by AI, and that question is answered task by task with the work scan from FTE TO AI. It is not personnel advice and not grounds for a dismissal decision; for the latter, separate statutory requirements apply.
Recurring revenue does not stand apart from the rest. Freed-up capacity resulting from this process must be visible in the reports a buyer is presented with, which connects to the question why management information weighs differently now that AI takes over work. The process itself must also be reproducible without relying on chance, a theme explored further on the page about why processes and systems weigh differently now that AI takes over work. And the question of whether that revenue depends on one person or one supplier directly touches the owner-dependency index addressed on the page about why team and succession weigh differently now that AI takes over work.
How these eight drivers together form the final price is described on the page how the value of a company is determined.
The question of whether your recurring revenue relies on people or on a process can often already be partly answered with a short look at your own records. For a broader picture, a free value check is available: eight short questions, one per value driver, resulting in which driver is currently weighing most heavily on your price. The full value scan, with maturity scores per driver and a two-year calendar toward the exit moment, is under construction.