Activating Employee AI Learning

With your AI council chartered, your galvanizers recruited, and your North Star metric identified, the next question becomes: how do you actually move that metric? Asana's answer was a learning and measurement strategy built on employee agency — giving people multiple on-ramps to AI skill-building rather than mandating a single path, then using research-backed surveys to see what was actually shifting beneath the surface. As Lisa Ann Logan described, this approach generated both momentum and the data to steer it.

You'll recall that Asana's Head of People, Anna Binder, set the stage at the annual kickoff with two messages: AI won't replace your job, but "it might start to be replaced by someone who knows how to use AI for more impact" — and "the number one thing that you can do for your career at Asana or anywhere else is to invest in learning AI." The team then had to deliver on that promise. Their internal learning conference, GrowthCon, added a dedicated AI track built with council members and internal experts — and it earned the highest NPS scores of any track offered.

Building on that momentum, Asana opened individual career growth budgets for AI learning. At a point when the company "wasn't yet ready to commit to an enterprise license for any one LLM," this let each employee choose their own entry point — LLM subscriptions, workshops, books, conferences — based on where they were in their journey. Roughly 20% of employees used their budgets for AI learning, which also gave the People team valuable signal on what kinds of learning employees found valuable without imposing a top-down curriculum. As Logan put it, "it put them in the driver's seat."

Employee agency drives enthusiasm, but without measurement you're flying blind. Asana launched a baseline survey in March measuring three dimensions: how useful employees found AI day-to-day (the council's North Star), self-assessed competency levels (novice through expert), and overall sentiment — generating over 3,600 comments that the team naturally analyzed with AI. By the November follow-up, results showed an 18 percentage-point jump from the novice and fundamental tiers into intermediate, and a 15-point jump into advanced and expert — pushing nearly 30% of the company into those top tiers.

What made this measurement strategy distinctive were two external research partnerships. Asana developed a five-stage AI maturity framework with Anthropic, benchmarking U.S. knowledge workers at 2.3 on the scale — early in what Logan called the "AI activation" stage (stage two of five). Asana itself scored higher, "growing out of the AI experimentation into AI scaling." Separately, an AI mindsets framework built with Carol Dweck's lab at Stanford segmented employees along an enthusiasm-to-skepticism spectrum — and surfaced a genuine surprise. Technical functions like engineering, product, and design "were actually super skeptical and we thought they'd be really enthusiastic." Logan's explanation was telling: "They're the ones building it. They've used it the most. They've gotten the most hands on there. Where are the limitations?" Without that function-level segmentation, Asana would have designed activation strategies based on assumptions that were exactly wrong for their most technically capable teams.