AI in Practice: The 5 Managers Test

Sample: 755 senior mid-market executives

Ready to take the 5 Managers Test? It’s pretty simple tbh. Here’s the question as we asked it in the survey:

“If we asked 5 managers to name your top 3 AI priorities, how many would agree?”

Now pick a number between 1-5.

Most AI programmes fail the five manager test. If four of them give you the same answer, you’re in rare company. In our study, fewer than one in four believe that would happen in their organisation.
That number matters more than almost anything else we measured.

Organisations where leadership aligns on priorities report positive AI ROI at nearly three times the rate of those where alignment breaks down – 79% versus 23%. We tested governance maturity, investment drivers, adoption stage, and various angles of infrastructure readiness, but the five-manager test outperformed them all as a predictor of whether AI is actually delivering.

It’s measuring whether the people accountable for outcomes are operating from the same map. GTM teams are the worst offenders (only 16% believe their peers would align), but the problem cuts across every function we surveyed.

The painful part is that most organisations have tried to get everyone marching to the same tune. They’ve sat in the all-hands where the AI strategy was presented and contributed to the roadmap. But then many leaders walk back to their desks and carry on with a subtly (or not so subtly) different version of what they think they should be doing.

At workforceLIVE on the 6th May, we’re going to dig into what it actually takes to close that gap in practice, from people who’ve done it. Because the data is clear: the technology usually isn’t the bottleneck – the real challenge is enrolling leaders and their teams to believe and buy-in.