An installation company lives on work that is difficult to capture in a system: a technician who assesses a breakdown on site, an estimator who adjusts a quote for a building that is not on the drawings, a planner who has to align three crews and a supplier at the same time. Alongside this is administrative work that can be captured: work orders, time registration, material orders, actual-cost calculations, maintenance contracts that require action at fixed moments. Both types of work together determine the cost price, but a buyer does not look at the hours on the work order. He looks at what keeps running when the current staffing changes.
The circumstances that drive the outcome are well known in the sector: an emergency repair service that revolves around one fixed person, a cost calculation that lives in the head of a single employee instead of in a template, a maintenance portfolio that is kept on paper or in a separate Excel file. None of this is wrong. It only determines how a buyer assesses the value.
AI is taking over tasks in parts of this work, not as a future scenario but as something that is already running in some installation companies and not yet in others. Three categories run through the entire company.
The first part is work that a system largely handles itself: reading work orders, linking material usage to a project, flagging when a maintenance contract is due for renewal. This is administrative work with a fixed structure, and this is where the largest share of freed-up hours lies.
The second part involves oversight: a system proposes a schedule, a draft quote, a priority order for the emergency repair service, and a person approves or rejects it, with reason. The quality of that oversight determines whether things go faster or just differently.
The third part remains human work: the technician who assesses on site what a drawing does not show, the conversation with a customer about a solution that does not fit the standard quote, the judgment of an experienced planner when two urgent jobs come in at the same time. This work does not change because a system cannot handle it, but because it revolves around judgment made on the spot.
Why it already works in one company and not in another usually does not lie in the technology. It lies in how the work is recorded. A company where cost calculation, planning and actual-cost calculation sit in a shared system can easily have those first two categories taken over. A company where that knowledge sits in people's heads and separate files must first organize it before there is anything to automate.
A buyer does not buy the number of technicians on the payroll, but what keeps functioning without the current staffing. Work that AI takes over therefore does not affect one driver, but the weight of all eight.
Owner dependency comes under the most pressure: if cost calculation, planning and customer contact rest on AI-supported processes rather than on one person, that weighs more heavily in the price than a company where that knowledge exists only in the head of the owner or a key figure. That is precisely why how owner dependency drags down the sale price and what can be done about it has a separate place in the value scan.
But profit quality also changes in character. A saving that arises because one supplier offers a favorable contract counts differently from a saving that is embedded in the process itself — for example because planning and actual-cost calculation structurally cost fewer hours, regardless of who the supplier is. A buyer values that second form more highly, because it is transferable.
The remaining drivers — customer concentration, growth scalability, contract structure, market position, system maturity and team depth — shift along as recorded work takes the place of work that depends on one person. A maintenance portfolio monitored through a system instead of through a planner's memory is, for a buyer, a different kind of asset than the same portfolio without that recording.
This does not happen in isolation from how the installation sector relates to other sectors. Anyone who wants to see how comparable shifts play out at care institutions where scheduling pressure and record-keeping determine value, at wholesalers where inventory and order processing form the core or at transport companies where planning and driver dependency raise comparable questions, will recognize the same pattern: recorded work counts more heavily than work that depends on a person.
On what AI does to staffing, a separate note applies. Whatever consequences an employer attaches to this for personnel falls under their own statutory requirements regarding dismissal and employee representation; that is not part of this page and not a topic on which advice is given here.
The underlying question — which work in this specific company can genuinely be taken over by AI, and which part remains human work — is answered task by task with the work scan from FTE TO AI, separate from what that means for staffing. For those who want to look more broadly at the path toward an exit, the steps to make a company sale-ready before a sales process begins provides a picture of what buyers test in practice.
Those who want to know which of the eight drivers is putting the most pressure on their own company's price today can fill in the free value check: eight short questions, one per driver, resulting in an initial picture of where the greatest pressure lies. The full value scan — with a maturity score per driver, the owner dependency index and a two-year calendar toward the exit moment — is under construction.