What does it actually takes to succeed at AI transformation at an enterprise level?
BCG found that 70% of it comes down to people and process. Just 10% is the algorithm and only 20% is the technology and data.
The lion's share — the part that determines whether transformation actually lands — has nothing to do with tools.
Every bit of it is about how you lead under uncertainty, how you communicate your value, and how you make the case for what you bring when the ground is moving.
Here's the counterintuitive part — and I need to say it plainly because everything else in this newsletter builds on it:
The AI problem for senior leaders is not about learning the tools.
Only 5% of the executives responsible for AI strategy use AI on a daily basis — compared to 57% of their technical teams. Yet tool proficiency isn't what's determining who advances and who gets sidelined. The anxiety running through senior leadership right now isn't "am I using the right platform?"
It's something structural underneath. Something the technology is exposing rather than causing:
The institution used to make the case for my value. Now it's changing. And I'm realizing I've never had to make that case myself.
That's the first part of what's at stake.
For most of a long career, the institution handled this quietly. The title said what you were worth. The scope of the role said where your judgment was needed. The org chart made the argument so you didn't have to.
What AI is disrupting isn't your expertise. It's the frame.
When the frame dissolves — through a reorg, an AI rollout, a market that stops needing the old category — you need language for what you bring. Your own. From the inside.
But that's only half of what makes you legible now.
The second half is this: you need a relationship to AI that isn't defined by tools.
Not a course. Not a certification. A thesis — a point of view on what this technology is for in your domain, what it should be directed toward, and what it cannot replace. That judgment is what separates the leader who is directing AI from the one who is being directed by it.
The thesis doesn't come from reading about AI. It comes from contact with it. From the moment — and there will be a moment — when something that should have taken three days takes forty minutes, and you feel the ground shift. That experience is not transferable. Someone describing it to you is not the same thing.
You need your own aha moments. And you need to know what you bring to the other side of them.
Two things. Not one.
Know what's irreplaceably yours — the judgment, the pattern recognition, the thing you've built across every role that was never the institution's to claim. And develop a real point of view on AI — not performed fluency, but an earned thesis about what it means for the work you do.
Together, those two things are what make you legible. To yourself first. And then to every room you walk into.