In many companies, contact with the most important customers runs through a small number of people: the owner themselves, an account manager who has been there for years, or an inside sales employee who knows who needs what and when. That works, until the moment a buyer looks at the company. A buyer will then wonder what remains of the revenue if that one person leaves. If the answer is unclear, that gets discounted in the price, even before anything else has been discussed.
This dependency usually did not arise on purpose. It is the result of efficiency: for years it was faster to just call than to document what was said. Knowledge about preferences, sensitivities and history stayed in people's heads because documenting it takes time that no one had. The company grew, the relationship grew along with it, and no one ever took the moment to transfer that to a system instead of to a colleague.
A buyer does not buy the number of people working there, but what keeps running without those people. For customer relationships, that means: does revenue continue if today's contact person is gone tomorrow. That affects not only the owner dependency index, but also earnings quality. A margin that runs on the goodwill of one relationship manager is valued differently than a margin that is embedded in the process and the systems. The same applies on the other side of the chain: how your dependency on suppliers has built up counts just as much for a buyer as the customer side.
What AI changes here is not that relationships become superfluous. It is that parts of maintaining a relationship no longer depend on a single mind. Roughly three categories run through this work.
The first part AI can take over: summarizing contact history, spotting patterns in ordering behavior, reminding a customer that a contract is expiring, or preparing a draft email based on previous correspondence. This is work that used to live in the account manager's head and can now sit in a system that everyone can consult.
The second part is work with oversight: AI proposes an approach for a difficult conversation, flags that a customer is ordering less than usual, or drafts a response to a complaint — a human assesses whether that fits this customer and approves or rejects it, with reason. Here the person does not disappear from the relationship, but the memory of the relationship no longer depends solely on that person.
The third part remains human work: the conversation itself, sensing the tone, building trust with a new customer. No system takes that over, and that is not the goal either.
Whether a company has already organized this way does not depend on the sector but on whether someone ever took the time to document what was in people's heads. Companies where that has happened can have a relationship manager step out without the customer noticing. Companies where that has not happened only notice it the moment someone leaves — or the moment a buyer asks about it.
Making it transferable does not mean that contact disappears or that a customer suddenly talks to a system instead of a person. It means that the information that now lives only in someone's head is also stored elsewhere: who the customer is, what is going on, what has been agreed, when action is needed. Once that is documented, a second person can take over the conversation without the customer noticing that anything has changed. That is the difference between a relationship that depends on a person and a relationship that depends on the company.
This also touches on who within the company is allowed to decide when the contact person is not there — a question that goes beyond customer contact alone and relates to how decision-making in the company is organized. And with commercial relationships, knowledge often comes along that is written down nowhere: which arguments work with which customer, who once dropped off and why. How that commercial knowledge is documented is also part of the same movement.
Where this touches on who continues to do which work within the company, the applicable statutory requirements apply; that is not part of this page.
Whether summarizing customer contact, flagging risks in a customer portfolio, or preparing conversations can be automated differs per company and per customer. That question — which work in this particular company can genuinely be taken over by AI — is answered per task with the work scan from FTE TO AI.
To see whether customer dependency is the driver putting the most pressure on your price today, or whether it lies elsewhere, there is the free value check: eight short questions, one per value driver, with an initial picture of where the pressure lies. For a broader picture of owner dependency in general, there is a separate deep dive on how you as an owner become less indispensable, and for those who want to look more broadly at what makes a company sale-ready, there is an overview of the steps toward a sale-ready company. The full value scan, with all eight drivers, evidence per maturity score and a two-year calendar toward the exit moment, is under construction.