The board meeting question has changed. A year ago, it was: “Do we need an AI strategy?” Today it is: “Why isn’t ours producing results?”
Most companies have already spent money on AI tools and licenses. Many have run AI pilots. Some have even hired a Head of AI. And most of them, if they are honest, cannot point to a meaningful return on any of it.
So what is actually going on?
The technology is ready. The organization around it is TBD.
When AI programs underperform, the default move is to look at the tools. Different model. Different platform. More licenses. Better integration. This is seldom the issue.
The companies getting real value from AI are using roughly the same tools as everyone else.
What separates them is everything around the tools: how teams are structured, how work is defined before anyone opens an AI assistant, how adoption spreads beyond the few people who figured it out on their own, and how results are measured relative to business outcomes.
Buying more software does not build that. Neither does hiring a Head of AI and hoping change follows.
Real transformation includes:
- changing how teams are composed
- defining the work before it reaches the engineers
- deciding which functions need to change, in what order, and who owns each part of that change
- building the measurement infrastructure to know whether it is working before the board asks.
These are decisions about governance, team structure, and how people actually work. They require leadership commitment, not just a bigger tech budget.
Most AI initiatives have the same four blind spots
No system behind the early enthusiasm.
In almost every company, a handful of people figure out how to use AI well and start seeing results. Everyone else watches and waits for someone to show them how. That moment rarely comes on its own. The knowledge stays with the early movers. The gap between them and everyone else widens.
No one to carry the change forward.
Top-down transformation runs into resistance. Bottom-up adoption stalls when the early adopters move on to other priorities. What is missing is a trained peer network: people inside the company who believe in it enough to lead others by example, not by mandate.
Governance that blocks instead of enables.
Most companies land in one of two places: Either they ban AI use until legal and policy teams catch up, or they let adoption run ahead of any framework and quietly pile up risk. Between those two is the right answer: a policy framework that enables responsible use from day one. Most companies do not build that spontaneously.
No way to measure success.
Without a system to track adoption and outcomes from the start, there is no way to know what is working, no way to show the board a return, and no way to make the case for continued investment.
The companies that build lasting advantage from AI are the ones that have built the organizational capability to keep improving as the technology changes.
You’re racing against the wrong clock
A lot of the conversation around AI right now is about not falling behind. But rushed, unstructured adoption does not produce lasting results.
The companies that get real, lasting value treat AI transformation as an organizational program, not a technology rollout.
For example, Infinum’s AI transformation consulting has a defined structure: a diagnostic phase to understand where things actually stand, an implementation phase where changes happen in the real work, and a handover phase where the capability becomes self-sustaining.
What matters is not the window to adopt AI. It is the window to build the organizational capability around it, before that capability becomes a baseline expectation and stops being a competitive edge.
