A buyer who requests your license file is not interested in the licenses themselves. He is interested in what those licenses say about how work in your company actually happens. A CRM subscription, an accounting package, an AI application that reads invoices or pre-sorts customer questions: each contract is a trace of a task that someone, or something, performs. Due diligence reads those traces because they are harder than what a management team says about the organization.
That makes the license list one of the most honest documents in the entire file. People can describe a process more optimistically than it actually is. A subscription that is billed monthly continues without a story.
Until recently, a license was mainly a cost item: something that weighed on the profit and loss statement and said little else about the quality of the work. That is shifting. Software that takes over tasks that used to be human work changes the meaning of a license line. Where a subscription used to be a tool alongside an employee, for a growing number of tasks it is now a replacement for that employee, or an intermediate form in which the system does the work and a human approves or rejects it with reason.
Three situations run together, and a due diligence tries to distinguish between them:
This differs greatly per company, and not because it is an industry that is ahead of or behind. It differs according to where the owner or team leader has personally put in time to actually transfer a task to a system, versus where a license has been purchased but the process has not really been redesigned around it. Two companies with identical software can therefore have a completely different reality behind the same line on the invoice.
If a task is actually done by software, that changes what a buyer is buying. He is then not buying the number of people currently on the payroll, but what keeps running when those people are no longer there. That is a different question than how many FTEs a company counts, and it directly touches on does a buyer buy your people or your way of working.
It also affects earnings quality. A saving that arises because one specific supplier temporarily performs a task cheaply is something different from a saving that is built into how the process itself is designed. The first disappears as soon as the supplier raises the price or stops; the second continues to exist under a next owner. Due diligence investigates which of the two it is, and that is exactly what what is normalization of profit at a company that uses AI is about.
There is also a third layer: owner dependency. If a task still runs on a person, and that person happens to be the owner who manages the license or assesses the exceptions, then the dependency has not been removed by the software, only relocated. A buyer wants to know whether the process gives the same outcome without that specific person.
This reading has limits that are worth mentioning. A license file shows what has been purchased, not what is actually being used. A subscription can be active while the underlying process still runs largely manually, or conversely: a task can already have been largely taken over by software without this being visible in a separate contract, because it is part of a broader license.
In addition, the presence of AI functionality in a package says nothing about the degree of oversight that is actually still maintained over it in practice. A task that "is done by AI" may in reality still be fully checked by an employee, which makes the effect on the valuation different than when that check is limited to exceptions. An outcome that only looks at license costs, without this context, says little.
This topic also touches on personnel decisions, for which separate legal requirements apply; these are not addressed here.
The question that due diligence tries to answer through the license list — which work in this company is truly done by AI, and which work only does so on paper — is mapped per task with the work scan from FTE TO AI.
This is also where recurring revenue and contract structure play a role: a saving that is embedded in how the company works weighs differently than a saving that depends on a cancellable subscription, just as what is recurring revenue worth in a sale shows that the form of an income stream counts just as heavily as the amount. And because a buyer often ties part of the price to performance after the transfer, it is useful to know how that link works, described in what is an earn-out and why does a buyer propose one.
A good starting point is gaining insight into which of the eight value drivers is putting the most pressure on the price today. The free value check consists of eight short questions, one per driver, and gives a picture of where the greatest pressure lies. The full value scan, with a maturity score per driver, evidence, the owner dependency index and a two-year calendar toward the exit moment, is under construction.