Reflection — Fewer AI projects, more value

The companies that get the most from AI don't do more, they do less: according to BCG, top performers prioritize an average of 3.5 use cases versus 6.1 for the rest — and expect 2.1 times greater return on investment. Spreading effort across twenty initiatives means truly succeeding at none.

The action: pick at most three high-stakes use cases, fund them all the way to scale, and say no to the rest for now.

Source: BCG, “From Potential to Profit: Closing the AI Impact Gap” (AI Radar), January 2025.