Your AI Rollout Isn't Failing at the Top. It's Failing One Layer Down.

Prompted by “Your AI Rollout Isn't Failing at the Top. It's Failing One Layer Down.”, Harvard Business Review (Dr. Gleb Tsipursky).

Executives love announcing an AI strategy. What actually determines whether anything changes is whether the people translating that strategy into daily work believe in it.

HBR's Dr. Gleb Tsipursky, drawing on his own research into AI adoption psychology, argues most AI rollouts fail one layer below the executive team — with middle managers, not the C-suite. He splits managers into five types, from outright skeptics to what he calls "AI catalysts," and the point is that each type responds to a completely different lever: some need evidence, some need incentives, some just need permission and support. Roll out one message to all of them and you'll convert the catalysts and lose everyone else.

What can a smaller company do differently?

At a leaner company, you probably don't have five layers of management to diagnose — you might have two or three people between you and the front line. That's actually your advantage: you can have this conversation with each of them directly, in a week, instead of running a company-wide survey. Most CEOs never do it anyway, because they assume "we're using AI now" is a decision, not something each manager has to personally believe is worth the risk.

Test it out: sit down with each person who manages people under you and ask them plainly where they land — are they actually experimenting with AI in their part of the business, waiting to see how it goes, or quietly skeptical it's worth the disruption? Don't guess — ask, and use different conversations for what each one actually needs to move.

This is exactly the kind of layer-by-layer work we do in Team Coaching — getting a leadership team genuinely aligned on something before it becomes a company-wide mandate.

Originally posted on LinkedIn ↗