Kennisbank

Value of an IT Company Now That AI Is Taking Over Work

Where the hours in the IT sector go

An IT company rarely sells only software. The largest share of hours goes to work around the product: implementation at the client, handling tickets, reviewing code, testing, maintaining documentation, writing quotes, planning projects. That work is partly standardized and partly dependent on who carries it out — a senior developer who has been in the codebase for ten years, an account manager who has known the client for years. A buyer looks at what keeps running if that person is no longer there, and in the IT sector that is more often an open question than owners think.

The shift that is already visible

AI is not taking over the same work everywhere in the IT sector, and not at the same pace everywhere. Three categories run through almost every part of a business:

1. Work that AI can take over: first-line support, generating standard code, drafting test scripts, summarizing documentation. 2. Work where AI makes a proposal and a human approves or rejects it with reasoning: code reviews, security assessments, architecture choices, tailored quotes. 3. Work that remains human work: client relationships at management level, complex problem diagnosis on unique systems, negotiating contracts.

In one company, the first category already runs largely through AI-supported tools, with a human checking the outcome. In another company, exactly the same task is still done entirely by hand. The difference lies not in the sector but in how the work has been documented: is the process recorded in a system with clear steps, or does the knowledge sit in the head of one developer who has served the same client for fifteen years. Documented work is easier to transfer to a tool with human oversight; undocumented work stays with the person, with or without AI.

What this does to the value drivers

This shift does not affect one driver but the weight of all eight.

• Owner dependency changes the most. If an owner personally maintains the most important client contacts or is the only one who oversees the architecture, that is a risk a buyer weighs regardless of what AI can take over. As for the question of which work actually depends on the owner and which work is transferable, separate statutory requirements apply as soon as this touches on personnel decisions — that is not part of this analysis.
• Profit quality is read differently. Savings that arise because one supplier works more cheaply count differently than savings built into the process itself, for example because testing structurally takes fewer hours. A buyer wants to know whether the margin still holds if the supplier changes or the contract expires. How you substantiate that is described in an explanation of substantiating profit quality in a sale.
• Recurring revenue carries a different weight when part of the underlying work becomes cheaper. A support contract that generates the same revenue but costs fewer hours is, to a buyer, something different from the same contract with the same cost structure from five years ago. What recurring revenue is worth in that light is addressed in an explanation of what recurring revenue is worth in a sale.
• Scalability shifts along with it: a company that can serve more clients without a proportional increase in FTE capacity, because part of the support work runs through AI with human oversight, shows a different growth curve than a company where every new client costs a proportionally extra number of people.

Not only an IT question, but an IT-specific answer

The same eight drivers play a role in every sector, but the specifics differ. In financial services, the dependency often lies in licenses and compliance knowledge, as described in an overview of what determines value in financial services; in construction, it more often lies in the relationship with subcontractors and permit processes, worked out in an analysis of valuation in construction now that AI is taking over work. In the IT sector, the center of gravity lies in knowledge fixed in code, documentation, and client relationships — and that is precisely what makes the sector sensitive to the question of whether that knowledge has been made transferable or not.

What you can do with this today

The underlying question — which work in your company can genuinely be taken over by AI, which work needs human oversight, and which work remains human work — is answered per task by FTE TO AI's work scan, regardless of what you do with it afterward.

To see which of the eight value drivers is weighing most heavily on your price today, there is a free value check: eight short questions, one per driver, giving you a picture of where the greatest pressure lies. The full value scan, with a maturity score per driver, an owner-dependency index, and a two-year calendar toward the exit moment, is under construction.