Last month I shared the synthesis report from my summer listening tour. This month I want to pull one finding out of it, because it's the one I haven't been able to stop thinking about, and because nearly every conversation I've had since the report went out has circled back to it.
It came from the organization in my tour with the most mature adoption measurement — a company that tracks actual AI usage, not license counts. When they correlated manager behavior with team behavior, the relationship was essentially one to one. Teams whose managers adopted and used AI, adopted it. Teams whose managers didn't, didn't. Not mostly connected. In the words of the executive who found it: one hundred percent connected.
Gallup's numbers say the same thing at scale. Employees who strongly agree their manager champions AI use are eight times more likely to say AI has transformed how work gets done: 33 percent against 4 percent. Only about a third of employees say they have a manager like that.
Once you see it, you see it everywhere. Permission, pace, and what gets rewarded all flow through the direct manager. An employee's sense of whether AI experimentation is safe isn't calibrated to the CEO's town hall; it's calibrated to the person who writes their performance review. A manager who visibly experiments creates a current the whole team swims in. A manager who changes nothing casts a quiet veto that no adoption dashboard will ever show.
And yet look at how most rollouts treat this layer. The tool ships the same way to everyone: same license, same announcement, same training module. Managers — the single most consequential audience in the building — get handled as just another user group. In my tour I heard about manager-specific training that was built and then never assigned or announced. I heard about manager enablement for thousands of employees landing on a tiny team outside HR. And I heard managers themselves describe being loaded up with more and more, with no time attached.
This isn't only a pattern I've heard. In a Gallup survey of 102 Fortune 500 CHROs this spring, 99 percent said AI is important to their organization's strategy, and half said they are not confident in their managers' ability to guide employees on using it. Only 57 percent are providing AI training for managers, and a paltry 8 percent have tied AI adoption to any performance incentive. We are asking this layer to carry the change while changing almost nothing about their conditions.
Here's why this pattern matters for leaders right now: two years into heavy AI spending, the manager layer is the cheapest leverage most organizations have left. No new platform, no additional licenses. Just a redirection of enablement attention toward the subset of people whose daily behavior sets the ceiling on everyone else's.
I go deeper on the mechanism, and the four moves real manager enablement requires, in this month's companion piece on Medium: The Manager Multiplier.
