AI: before the tool, the decision

In 2025, 16.4% of Italian companies with ten or more employees used AI, double the previous year, yet about 95% of generative AI pilot projects produce no measurable effect on the bottom line. Eight decision processes and two axes, criticality and data readiness, show where to begin: first the decision, then the data, then the model. The tool comes last.
VEDRAI OBSERVATORY
AI: before the tool, the decision
Where and how to bring AI into the company: eight decision processes, two axes, one map.

Companies pick AI starting from the available tool, not the decision to improve, and end up with individual use and pilots that do not scale. Vedrai Observatory read eight processes along two axes: criticality and data readiness. The place to start is where a critical decision meets data that is already ready; elsewhere, a measure comes first.

KEY INSIGHTS

  • The useful question is not which tool to adopt, but which decision costs the most.
  • On numerical decisions, generative AI answers probabilistically, not by calculation.
  • Start from a critical decision with ready data; otherwise, build a measure first.