A buyer does not pay for the number of people on the payroll. They pay for what keeps running if one of those people disappears tomorrow. That has always been true, but the question takes on a different weight now that part of the work no longer sits with a person, but in software, procedures, or a combination of both. Team and succession used to revolve mainly around the question of whether there is a second person who can take over the owner's work. That question remains, but a second layer is added: which part of the work no longer needs to be taken over by a person at all, because a system already does it, or largely does it.
These two layers together determine how heavily this driver weighs on the price. A business in which the owner is the only one who maintains the key customer relationships, the only one who knows how the primary process works, and the only one who makes the decisions, is a different risk to a buyer than a business in which that knowledge and activity is spread across people and systems.
Work that AI takes over does not do so in the same form or at the same pace everywhere. Within team and succession, three categories run alongside one another. Part of the work can be taken over entirely by a system: fixed reporting, standard correspondence, scheduling according to fixed rules. Part happens with oversight: an employee approves or rejects, with a reason, before something proceeds. And part remains human work: negotiating with a hesitant customer, making a decision that does not fit a procedure, managing someone who can no longer see the way forward themselves.
The difference between businesses does not lie in whether AI is present, but in how much of the second and first type of work has already been identified and organised. In one business this has been examined and documented: which tasks a system can handle, which an employee carries out with oversight, and which remain with a person. In another business, everything still hinges on what the owner or a key employee carries in their head, without anyone having worked out what part of that is actually transferable to a system. That difference is measurable, and it is exactly what a buyer asks about.
A buyer rarely asks directly whether the business depends on you. They infer it from a number of signals. Whether processes are documented or known only to one person. Whether customer relationships belong to the business or to an individual. Whether decisions can be made somewhere other than at your desk. Whether there is someone who, with or without AI support, could take over tomorrow what you do today.
The owner dependency index brings this together into a single picture: how much of the revenue, customer contacts and operational decisions literally hang on the owner, and how much of that is transferable with reasonable effort, to a person, a team or a system. High dependency depresses the price, regardless of how good the figures are otherwise, because the buyer is buying, from day one, a risk they will have to resolve themselves.
The first step is an honest inventory of what happens if you are unavailable for two weeks. Not as a thought experiment, but on paper: which decisions stall, which customers no one else calls, which report fails to appear. That list is the raw owner dependency.
The second step is to look, per task on that list, at what transfer already exists: a colleague who knows the process, a documented procedure, a system that handles part of it automatically. Here team and succession directly touches the other value drivers. How processes and systems themselves are measured in a business where AI is contributing is a separate question, but the answer to it largely determines how high owner dependency turns out to be here. A process that exists only in the owner's head is a different risk than a process that is documented and partly carried out by a system.
The third step is distinguishing between what AI can take over, what AI can do with oversight, and what a person must keep doing. That distinction is precisely what the work scan from FTE TO AI answers: for each task in the business, it is established whether AI can take it over, can partly carry it out with oversight, or whether it remains human work, which makes owner dependency immediately concrete and transferable instead of a feeling.
Lowering owner dependency is not a matter of working harder, but of documenting, distributing and, where possible, automating what now sits with one person. Whatever part of this becomes a personnel matter falls outside what can be described here; decisions about who continues to do which work are for the employer, subject to their own applicable legal requirements. What this does concern is the factual question of where work sits and what part of it is transferable, separate from who ultimately carries out that work.
This driver does not operate in isolation from the rest. Lower owner dependency also affects how a buyer looks at the composition and origin of earnings quality and at the substantiation required for that, because results that do not hang on one person are generally valued more highly than results that come under pressure when that person leaves.
The free value check consists of eight short questions, one per driver, and gives a first picture of which driver is weighing most heavily on your price today. For team and succession, this is often the most honest starting point: a few questions about what happens if you are unavailable immediately reveal where the greatest dependency lies. The full value scan, with the owner dependency index and the two-year calendar toward the exit moment, is under construction.