NewAI Adoption Index

Writing / Strategy, Adoption to value

Everyone is adopting AI. Almost no company is getting more valuable.

People are faster. The company is worth the same. Here is why those two facts sit side by side, and what closes the gap.

Walk into almost any company and you will see the same thing. A marketer drafts faster. An analyst clears a backlog. An engineer ships more. Leadership buys licenses, runs a training, and watches usage climb.

None of that is a mistake. The instinct is right and the gains are real. A company that is not doing this is already behind.

But look at where the gains land. BCG's 2025 research on the AI value gap found that only about 5% of companies capture value from AI at scale. Nearly everyone is adopting. Almost no one is compounding.

The gain stays with the person

When one person gets faster with AI, the improvement lives in that person. It lives in their prompts, their setup, their sense of what to ask. The company does not own any of it. If they leave, it leaves with them.

So the person gets faster while the company grows no more capable. Those are different things, and most dashboards only measure the first.

What piles up instead

There is a second effect that leaders tend to miss. Each person adopts AI in their own way, with their own tools and their own habits. None of it was built to connect to anyone else's.

Across a hundred people, that produces a hundred private versions of how to do the work. Each one makes sense by itself. Together they are a surface the rest of the company cannot use.

Something always compounds. With individual adoption, the thing that compounds is fragmentation.

What closes the gap

Value compounds when capability belongs to the company and not only to whoever works there this year. That takes a foundation individual adoption cannot produce:

  • Knowledge the company holds, instead of context that lives in people's heads.
  • Orchestration that moves work cleanly between people and AI.
  • Governance so there is one trusted answer, and each person and agent sees only what they should.
  • Measurement so you can prove what changed.

This is built, the way you built your finance systems and your data platform. It is not trained into individuals.

This is a bet on people

Moving routine work to AI is not about needing fewer people. It is about what your people do with the time: set direction, use judgment, serve customers, make the calls no model can own.

The choice is simple to state. Let your capability keep living in individuals who come and go, or build something the company owns and improves every cycle.

Next in this series: Me, us, in sync.

Put AI to work across your company.