If we want to talk about enterprise AI adoption, we need to stop starting with strategy

Strategy is how organisations talk about change.

Behaviour is how change actually happens.

Most AI conversations begin with grand ambition. Transformation. Reinvention.

But don't we need to start with observed behaviour?

Of how people are already using these tools.

Every day. Unsupervised. Without ‘permission’. In ways they wont readily admit to.

Until recently, that data was mostly anecdotal. Then ChatGPT released its version of Unwrapped.

Like Spotify (but more like Black Mirror) it's a simple but interesting summary of how we had actually used ChatGPT over the year.

What’s most interesting was how each user seems to have been assigned an archetype.

That archetype was described. And, crucially, it was shown as a percentage of the total user base.

In other words: behavioural patterns at population scale.

So I asked ChatGPT a simple follow-up question:

If this is how individuals use AI today, what are the dominant behaviours and how common are they?

Here’s what came back.

The dominant AI usage archetypes (and how common they are)

Archetype	Share
Executor	38%
Translator	22%
Sense-Maker	18%
Reassurer	12%
Explorer	7%
Strategist	3%

Most usage clusters around execution, translation, and clarification.

Very little of it is strategic.

Almost none of it is transformational.

Not a surprise. Not really.

What this actually tells us

The majority of users are not redesigning operating models… They are:

• Writing emails faster

• Summarising things they didn’t want to read

• Checking tone before sending something risky

AI is spreading because it reduces cognitive, social and political friction.

Not because it unlocks bold new futures.

Not yet.

And, yet...

Most enterprise AI programmes seem to be designed for the Strategist archetype. A minority who think in terms of systems and second-order effects.

The problem is that Strategists represent about 3% of actual users. They are the edge case.

And if we design for the edge case, adoption will stall. Or fail. Every time.

Yet. This small minority do extract disproportionate value.

They use AI to remove steps, collapse workflows, and eliminate work that no longer needs to exist. They rethink how value is created.

They don’t prompt any better than anyone else. They think differently.

Because AI maturity is behavioural not technical.

And, people don’t start out as Strategists.

They move through: execution - sense-making - translation - confidence-building…

Only then does systems thinking become possible.

The big reveal?

AI in the enterprise is revealing a new future of work while exposing the current one.

A world of overloaded people, fragile confidence, misaligned incentives, task switching and constant translation between silos.

That’s why adoption is happening bottom-up.

And why the real opportunity isn’t tools.

It’s designing organisations for how people actually think and work.

Not the fictional competence we role-play at strategy off-sites to comfort ourselves.