A buyer does not count heads. A buyer asks itself one question: what keeps running if a key person leaves tomorrow. The answer to that question determines whether a company is valued on last year's profit, or somewhat lower because part of that profit depends on people who are not sold along with the business.
This is not a new way of looking at things. Due diligence has always asked about transferability. What is changing is the way companies organise that transferability. Work that used to exist only in an employee's head is, in a growing number of companies, being captured in a system that carries it out or checks it. That changes not only how much work there is, but who or what carries the work.
AI is taking over work. Not everywhere, not all at once, and not as a promise for later — in parts, already today, across three categories that run through almost every process. Some tasks a system can carry out entirely on its own. Other tasks are partly taken over, with a person approving or rejecting and giving a reason for it. And a third category remains human work, because it requires judgement, relationships or context that a system does not have.
The split between these three categories varies strongly per company, even within the same sector. The difference does not lie in the sector but in how the work is organised: is the process described and repeatable, or does the knowledge live only in the head of the person who has been doing it for years. Companies where processes have been documented see more quickly which tasks lend themselves to being taken over. Companies where the work is passed on verbally have something else to do first before AI can take anything over there: making the work visible.
Once work shifts from a person to a system with oversight, the weight of every value driver a buyer takes into account changes. Owner dependency the hardest: if the owner is the only one who knows which exceptions a customer gets, no amount of automation counts that knowledge — the dependency remains. But profit quality also changes. A saving that shows up in today's profit because one external supplier automatically carries out a task is different from a saving that sits within the company's own process and will still work tomorrow if that supplier stops or raises its price. A buyer who does not see that difference is calculating themselves richer or poorer without knowing it. What happens to value when the work depends on a supplier describes exactly that distinction.
A maturity score per value driver is an estimate, not a measurement. It is based on evidence — documented processes, systems that carry out tasks, documentation that makes handover possible — but evidence is not equally easy to find everywhere. For a task a system carries out entirely, that evidence is often simple to show: a log, a report, a result. For a task that is partly taken over with human oversight, the evidence is softer: someone approves or rejects, but the reason for it is not always fixed on paper. There the estimate is less certain, and that is not hidden behind a score that looks more certain than it is.
A score also says nothing without the question of who should believe it. A maturity score that rests only on internal impressions does not convince a buyer. How do you substantiate a saving that AI has delivered shows which evidence holds up in due diligence and which evidence only convinces internally. And without recording exactly what AI does in the company — which task, which system, which oversight — there is little to fall back on; how do you record what AI does in your company covers that record-keeping.
The outcome of the scan mainly says something about where a buyer would start asking questions. It says nothing about what should happen with staff. If an outcome touches on the question of whether roles should disappear or change, separate statutory requirements apply, independent of this method.
Underneath all the drivers lies one question: which work in this company can genuinely be taken over by AI, and which work cannot. That is the question the work scan from FTE TO AI answers per task, using the three categories as its starting point.
Making a company sale-ready is not a single intervention but a series of choices over two years, where each choice makes a driver a fraction stronger. How do I make my company sale-ready sets those choices out on a timeline. Owner dependency is within that often the driver that weighs most heavily, and also the slowest to reduce — transferring knowledge and letting systems take part in the work takes time. How do I become less dependent as an owner describes where that reduction usually begins.
The question of whether a buyer will one day be buying people or a way of working cannot be answered in a single conversation. A first impression can be. The free value check consists of eight short questions, one per value driver, and shows which driver is weighing most heavily on the company's price today. The full value scan, with evidence per driver and a two-year calendar toward the exit moment, is under construction.