A buyer at the table asks different questions than an advisor who stops by for a periodic check. An advisor asks how things are going. A buyer asks what remains once he himself is at the helm, and with which people, which systems and which dependencies. At that moment, what counts is no longer how many FTE are on the payroll, but what still keeps running without part of that input. If in your company work is already partly done by AI — with human oversight that approves or rejects, or sometimes without oversight because the task has been fully taken over — that changes the calculation the buyer makes, even if you had not yet made that calculation yourself.
That is the difference with previous sale moments. Five years ago, the conversation was about headcount and overhead costs. Now the conversation is about which part of the work hangs on a person, which part on a system, and which part on a supplier who in turn deploys AI. That third category is new, and it is precisely this category that changes earnings quality: a saving you realize because one external party takes over a task weighs differently than a saving that is anchored in your own process. The buyer wants to know which of the two he is taking over.
The takeover of work by AI does not proceed at the same pace everywhere. In a company where reporting, planning or first-line customer contact already runs largely systematically, there is more work whose takeover is demonstrable: there is a process, there is a log, there is a result that can be traced back to the system rather than to an individual. In a company where the same tasks still rest with one person — often the owner themselves — nothing is demonstrable, even if that person is in fact already using AI tools. The difference lies not in the technology but in whether the work has been detached from the person currently performing it.
That directly affects the owner-dependence index. If a buyer determines that a substantial part of the revenue, the customer relationships or the operational decisions is carried by you personally, every other value driver weighs more heavily against that dependence. Work that AI has taken over can lower that dependence — but only if the taken-over work is also documented separately from you, repeatable and transferable to whoever runs the company. Takeover by AI without transferability of the process does not change the index.
At the moment the buyer is at the table, you can establish which tasks are today demonstrably done by a system, which tasks run under oversight with a recorded reason for approval or rejection, and which tasks rest entirely with a person. That is a factual inventory, not a prediction. What you cannot establish — and what is also not promised by anyone working on this subject — is whether a task that today is still human work will be taken over tomorrow. That shift happens in parts and not everywhere at the same pace, and no one guarantees its tempo for your specific company.
Nor can you, on the basis of this inventory, make a statement about what should happen with your staff. Which work AI does or does not take over is a factual question; what an employer does about that on the side of its staff is a decision with its own statutory requirements, and different frameworks apply to that than to a valuation for a buyer.
What is within reach, however, is clarity on the question the buyer is actually asking: which work in this company can genuinely be taken over by AI, independent of who currently performs it. That question is answered per task with the work scan from FTE TO AI, with a score for each task in the category that fits — taken over, under oversight, or human work.
The way this question lands differs per sale scenario. With a full transfer, the emphasis lies on what keeps running without you; with a partial sale with the work that AI has taken over, what mainly counts is whether the remaining stake weighs as much as the part sold. With a transfer within the family with the work that AI takes over, a different dynamic plays out: the successor already knows the company, but the question of whether processes exist independently of you remains the same. Even apart from an imminent sale, it is worthwhile to know how this works if you do not yet want to sell but do want to know what the work that AI takes over means, because today's inventory is the same one a buyer will make later.
In addition, what is contractually fixed also counts: which contracts increase the value of a company and how customer concentration factors into a buyer's judgment, as explained in what customer concentration is and why a buyer factors it in, together with the degree of AI takeover determine how much weight a single dependence carries in the balance.
If there is a buyer at the table, the first useful step is not a fully revised valuation but a picture of where the pressure is coming from. The free value check consists of eight short questions, one per value driver, and gives a picture of which driver is putting the most pressure on your price today. The full value scan, with a maturity score per driver, the owner-dependence index and a two-year calendar toward the exit moment, is under construction.