Insights

What the numbers say, and what they mean

Notes and analyses on AI adoption. Grounded in research, data and analyses I have gone through carefully myself.

Analyses and notes

Analysis · AI pilots

95% of AI pilots fail. Here is why.

It is rarely the model that is the problem. MIT NANDA's study "The GenAI Divide" shows that almost all enterprise pilots strand without effect on the bottom line. Not because of resistance, but because no one has taken responsibility for the way forward: neither ownership, structure nor a plan for operations.

Analysis · Transformation

10/20/70: where the AI investment is decided

BCG's principle for digital transformations: 10% is about algorithms, 20% about technology and data, and 70% about people and processes. Most budgets are allocated in the opposite order. That is where the gap between investment and effect appears.

Analysis · Adoption & growth

The AI gain follows the depth, not the purchase

Data from Ramp Economics Lab and Revelio Labs suggests that companies with deep AI use grow markedly faster in headcount (+10.2%) than those that bought the tools without changing their ways of working. The gain follows the depth of the integration.

Series · Notes from reality

Notes from reality

A series of physical, handwritten notes on AI and people. All images are photographed, not generated. The series runs on LinkedIn.

Want the insights translated to your organisation?

Write a few lines about what you are facing, or start with the readiness test.

Send a message