Strava route recommendation engine: athlete-centric personalization

AI · 5 min read

Strava's route recommender now incorporates richer athlete intent signals — preferred distance, elevation, and social patterns — along with environmental data like bike lane availability and traffic heatmaps. The ranking favors routes that match historical pace and preferred surfaces while penalizing high-risk roads during peak hours. This results in more usable suggestions for training and leisure rides alike.

The UX allows quick toggles for intensity, scenic routes, and safety, and previews show estimated time, calories, and climb. Strava also surfaces community-driven notes (e.g., construction warnings, water points) which improve route reliability. Internally, the system blends graph-based pathfinding with learned user preferences to generate both familiar and exploratory options.

For product teams in health and navigation, prioritizing safety signals and making personalization adjustable are critical. Strava demonstrates how combining community data with explicit user intent yields recommendations that feel both relevant and trustworthy.