An interview with Strava CEO Michael Martin on Athlete Intelligence, the premium-only AI “coach” debuted in 2024 that has alienated experienced users, and more
Almost like human connection matters or something [embedded post] @jimmytwohands : Strava is funny because there are little segments called “bloopie sprint” or “twirly deer crossing” and you can check how fast you biked it compared to your last time and also see who went the fastest and it's always some tour de france motherfucker who blasted through at 40 mph Hazard / @hazardfromharvard : Been a long-time Garmin/Strava user and have to believe there's some value-accretive M&A possible in the space (possibly by Apple?) to enhance hardware and software features and user experience.
Context & Ripple Effects
Strava had already demonstrated a subscription business, with earlier reporting indicating strong revenue growth and millions of paid users; Athlete Intelligence tests whether AI can deepen that paid relationship without weakening the product’s appeal to its most committed athletes. Strava also expanded its fitness-platform reach through an Apple Fitness+ partnership, making the quality of its software experience more consequential alongside hardware ecosystems.
The interview surfaces a product-design tension: an AI feature intended to add premium value is drawing criticism from experienced users, whose expectations are shaped by the detailed activity data and comparison tools that define Strava’s core experience.
First-order effects
- Athlete Intelligence remains a premium-only feature, so paying users are the immediate audience for its value proposition—and experienced users dissatisfied with its coaching output are the immediate retention risk.
- Strava must weigh feedback from its most knowledgeable athletes against the subscription rationale for keeping AI coaching behind Premium; this is a product-positioning issue rather than simply an AI launch.
Second-order effects
- Fitness platforms and device ecosystems can compete by pairing activity data with coaching features that feel more tailored or controllable, raising the bar for Strava’s premium experience.
- Strava’s deeper Apple Fitness+ integration gives it another distribution surface, but also makes consistent workout interpretation and useful summaries more important to the combined user experience.
Third-order effects
- Consumer fitness AI is likely to be judged less by the presence of a chatbot-like coach than by whether it earns trust from highly engaged users with established training habits.
- The broader subscription opportunity may shift toward configurable, specialized coaching layers rather than one-size-fits-all AI features, especially where platforms already hold years of athlete data.
The trend: Fitness apps are turning proprietary activity data into paid AI coaching, but durable differentiation will depend on user trust and training relevance rather than AI branding alone.