A private-equity partner does not primarily look at whether AI can take over work. He looks at what that does to the valuation of the portfolio company and to the path toward exit. A cost structure that changes, changes a multiple. A competitor that restructures faster than the portfolio companies, changes a market position on which the investment memorandum was built. The question is not "can we do this", but "is the assumption on which we bought in or are going to sell still valid".
That makes the position different from that of a CEO or CFO. Where the question a ceo asks often concerns execution and management, and where the question a cfo asks revolves around margin and cost structure, the PE partner reasons back from the next transaction. For him, every shift in what a company can do with AI is, first and foremost, a question about price.
Here too, the three categories run through everything. In part of the tasks within a portfolio company, AI can take over the work; think of structured reporting, first-pass analyses, recurring control activities. In a larger part, the work is done partly by AI, with an employee who approves or rejects it and can substantiate that. And in part, it remains human work: client relationships, negotiation, judgment under uncertainty.
Exactly where that boundary lies depends on the company, not on the sector as a whole. Two portfolio companies in the same industry can show a very different outcome here, because one has already broken the task down into steps that AI can take over while the other still organizes the work as a single whole. That difference is measurable, not predictable on the basis of the sector clock alone.
A PE partner does not accept an answer based on a sector average or a trend line from a market report. A multiple is not revised on the basis of what AI does "on average" in an industry; it is revised on the basis of what this company, with these tasks and this workforce, can demonstrably take over or not. An assumption such as "our cost base is structural" is only tenable for him once he knows what part of that cost base consists of work that can now, partly or fully, be done by AI.
Nor does he accept an answer that already answers the question about personnel. Whether and how a company adjusts its workforce to a changed division of tasks is a decision for the employer, with its own legal requirements that decision must meet. The role of the PE partner here is limited to assessing the facts about the work itself, not to supplying arguments for a personnel decision.
The friction with management usually arises over timing. A CEO wants to embed a shift in the work before he brings it into the open: first let the organization move along, only then report on it. A PE partner wants to know whether the assumption underlying the valuation still holds at the moment he needs it, regardless of whether the internal adjustment is already ready for that. That difference in pace cannot be resolved by working harder; it is a difference in what the two roles see as their task. The CEO steers the organization, the PE partner steers the ownership of the organization.
The same friction plays out with the bank, when financing is at stake: a lender assesses an AI shift mainly on what it does to cash flow and collateral, as described at the test a banker applies to a changed cost structure. A PE partner and a banker thus ask different questions about the same shift, and a board that prepares only one answer gets stuck on the second question.
Even within the board itself, interests do not automatically align. A supervisory body must be able to test an AI decision without taking over the execution, and that places it in a different position than the investor who is looking at the next transaction; how that testing role relates to a shareholder with an exit horizon is described at the role of the supervisory board in AI decisions. For the PE partner, that distinction matters because a body that does not move along with the shift can cause a delay in the valuation that was not accounted for in the model.
Underneath many of these discussions lies a confusion that is not unique to investors: a plan holds up as long as the figures check out, a strategy fails as soon as an underlying assumption no longer holds, even if the figures do not yet show that. The difference between the two is set out at the page about the difference between a strategy and a plan, and anyone who wants to know how an assumption is tested for validity will find that at the explanation of what a strategic assumption is and how you know whether it still holds.
The underlying question — which work in this company can genuinely be taken over by AI — is answered per task by the FTE TO AI work scan.
A PE partner who wants to be sure whether the valuation of a portfolio company still rests on valid assumptions can start with a free assumption check: a short round in which you name the key assumptions and see, for each one, when it was last confirmed. The full strategic pressure test is under construction.
Stel uw vraag. Vaak zit de echte vraag een laag dieper — daar mag ik naar vragen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.