There's a pattern we see in almost every AI conversation with a leadership team. A tool shows up, the price looks reasonable, someone does the math on whether it pays for itself, and the conversation ends there.
The company approves the seat, the tool gets used a few times, and three months later someone asks whether it's still worth the subscription. Nobody's quite sure, so usage quietly tapers, and the €100 becomes the number everyone remembers when they hear the words "AI tools." That's the wrong conversation, and it's the one most Nordic companies are having right now.
Take one senior knowledge worker: an analyst, a lawyer, a consultant, any role where the output is thinking, writing, research, or judgment applied to the work of clients and internal stakeholders. Measure the ceiling of their capacity, meaning how many clients they can take on, how many matters, cases, accounts, or projects. That ceiling used to be a function of their hours, their focus, and the manual work surrounding the actual expertise, and the manual work is where the ceiling gets set.
When the tool is integrated properly into the way that person works, the manual work collapses while the expertise stays intact. Four hours of work takes forty minutes, five back-and-forth drafts collapses to one, and the constant switching between tools, file systems, and mental contexts now runs in parallel, driven by an assistant that holds the full context in one place.
What actually changes
- The capacity ceiling moves. A senior operator who handled ten matters comfortably can handle three to five times that, at the same quality and with the same focus. Not because they're working harder. Because the work around the expertise has collapsed.
- The ratio inverts. Expertise used to be 30 percent of the hours and manual work 70 percent. That flips. The senior person spends most of their day doing the thing only they can do.
- Decision speed compounds. Faster analysis, faster drafts, faster iteration means decisions that used to wait a week happen the same day. Over a quarter, this reshapes how the business moves.
- Throughput scales faster than payroll. The senior people in the business produce more. The leverage from each hour goes up. The margin difference is what the board actually sees on the P&L.
The trap most companies fall into
The €100 conversation is a symptom of a bigger mistake. Most companies treat AI tools as subscriptions to evaluate, not capabilities to integrate.
They approve the seat, tell the team to try it, and run it alongside an existing workflow that was built for the pre-AI capacity ceiling. The tool does small helpful things inside an unchanged process, and three months later the math still looks like €100 a month for a little bit of help.
That's the "done right" distinction. The uplift isn't in the tool, it's in the workflow that gets rebuilt around what the tool now makes possible. Companies that do the rebuild stop counting the €100, and the ones that skip it keep counting, then quietly conclude that AI tools are overpriced.
Where the real margin sits
Think about what a three-to-five times capacity increase on a senior operator actually means.
A law firm's senior associate handling three times the matters at the same quality, a strategy consultant shipping three times the recommendations without dropping depth, an in-house analyst covering three times the territory on the same team. Same hours, same people, different ceiling.
That's not a tool upgrade, it's a capability shift, and the companies whose senior people have made the shift are pulling away from competitors whose people haven't. The €100 rents the tool from a vendor, but the integration is what creates the value, and the integration is what gets built inside the company.
The uncomfortable question
If you looked at your company's AI tool subscriptions right now, what would you find?
If the honest answer is "a few seats, some light usage, and a general feeling that AI hasn't delivered," you're running the tool-evaluation conversation. The €100 is still being counted, because nothing around it has been rebuilt.
That's the trap. The work around the tool hasn't changed, so the tool can only deliver what an unchanged workflow allows it to deliver.
The companies that move first are the ones that stop asking whether €100 is worth it, and start asking what their senior people's ceiling would look like if the work was rebuilt around what's now possible.