In the high-stakes world of corporate digital transformation, there is a new favourite metric: the "AI Adoption Rate." It sounds scientific, it looks great on a PowerPoint slide, and it usually involves a very impressive-looking chart showing thousands of Copilot licenses deployed.

There's just one small, slightly awkward problem: having a license doesn't mean someone is actually using it. Let alone, using it in a way that creates any value. In fact, many Swiss organisations have discovered that while they've successfully "deployed" AI to thousands, a staggering majority of those employees haven't sent so much as a single AI-assisted message.

Welcome to the era of "Shelf-ware AI."

The "IT Project" Trap

Many organisations are treating AI adoption like a standard software rollout. It's the "Field of Dreams" approach to technology: if you buy the licenses, the productivity will come.

But AI isn't like Excel; you don't just "click" your way to success. Measuring adoption solely by tool usage numbers like license activations or daily active users is easy to collect, but it proves almost nothing about actual value. It's the equivalent of measuring a gym's success by how many people bought a membership, rather than how many people can now lift their own groceries without groaning.

When Metrics Meet Reality

Currently, organisations are fixated on adoption and usage metrics. They are tracking:

  • License activation rates: (Translation: "They logged in once")
  • Average interactions per week: (Translation: "They use it as their personal therapist")
  • Time spent in the tool: (Translation: "They are staring at the blinking cursor in confusion")

Meanwhile, only a tiny minority is looking at the "Behavior Signals" that actually matter, like whether employees are proactively sharing what they've built, or if AI is becoming a genuine habit.

The organisations that treat AI as an IT project will get an IT project. Those who treat it as a human transformation will get a transformed organisation.

The Missing Link: Human Transformation

Real success isn't found in a "usage intensity" dashboard. It's found in the "Missing Middle": the space where employees move from "prompt users" to "everyday AI" habits.

The problem isn't that people lack access to the tools. It's that no one has helped them build a relationship with those tools. Behavior change is slow, nonlinear, and deeply personal. A license does not a habit make.

How to Stop Measuring the Wrong Things

If you want to know if your AI investment is actually doing something besides making your IT department look busy, shift your focus:

  • Measure Behavior, Not Just Usage: Active licenses tell you nothing. Instead, ask: are people using AI proactively? Are they sharing their "citizen-developed" automations?
  • Look for Habit Formation: Is AI embedded in daily workflows and team OKRs?
  • Track Knowledge Sharing: Spontaneous AI evangelism within teams is a much better signal than a "pre-and-post confidence score" that was probably filled out during a lunch break.

In short: stop counting the "clicks" and start counting the "ah-ha!" moments. Because at the end of the day, an unused AI license is just a very expensive way to make your "Digital Transformation" slide look colorful.