Worklytics

Evan Franz, MBA

216% higher odds of trying AI. That's what researchers found when more than 25% of an engineer's broader peer group had recently used Copilot CLI, compared with engineers whose peers hadn't used it. Having a manager who used it was associated with 82% higher odds. The takeaway? AI adoption spreads through people, not just company-wide rollouts. Our Worklytics team recently explored this in a piece on finding hidden AI power users. And I think companies are overlooking some of their most valuable resources. The employee driving AI adoption may not be the person leadership nominated. It could be a newer hire quietly improving a workflow or someone in a low-adoption department who's figured out how to make AI genuinely useful. Traditional dashboards often miss them. LayerX research found that half of employees had 12 or fewer AI conversations, while the top 5% had 144 or more. Nearly 47% of enterprise AI conversations in that report happened through personal accounts. That makes identifying your real power users harder than checking a single vendor dashboard. But high usage alone isn't the goal. Someone generating 30 drafts isn't necessarily creating more value than someone using AI to make 1 better decision. The real opportunity is finding employees who are getting better results and helping others learn how. A few things I'd prioritize: 1. Look beyond team averages. Strong AI users can exist inside low-adoption departments. 2. Measure outcomes, not just activity. Find where deeper AI use connects to better work. 3. Make successful workflows visible. Practical examples from trusted peers are often more useful than generic training. 4. Give champions time and support. Then measure whether adoption and outcomes improve over the next 12 weeks. The research shows a strong association between peer usage and adoption, not proof that appointing champions causes the same effect. Still, the implication for leaders is significant. Your next wave of AI adoption may already be sitting inside your organization. The challenge isn't finding more people to tell about AI. It's finding the people who've figured out how to use it well and helping their knowledge spread. https://lnkd.in/er94N9_D