Strava’s new Performance Predictions may be the reality check you need

Strava’s fitness tracking platform just got smarter. The company has now launched Performance Predictions, a new feature designed to estimate personalized finish times and paces for athletes training for popular race distances, including 5K, 10K, half marathon, and marathon. In other words, it will tell you exactly how fast you’re likely to finish 13.1 miles based on your recent runs, not your New Year’s optimism.

Powered by a machine learning model, Performance Predictions analyzes over 100 data points, comparing your stats to similar athletes. It updates automatically after each run, adjusting even during rest periods to reflect your current performance and fitness status. The goal is to provide dynamic, data-driven predictions that help you train with greater intention. By tracking your recent runs and recovery, the tool helps you stay on top of your progress, align your training with your goals, and follow an effective plan all the way to race day. It takes some guesswork out of training and replaces wishful thinking with reality-based forecasting.

Of course, this tool is available exclusively to Strava members, meaning you’ll need a subscription to take advantage. If you fall into that camp, you can find Performance Predictions in the app’s You tab. If not, a membership costs $11.99 per month or $79.99 annually.

The launch of Performance Predictions is especially interesting in light of Strava’s recent acquisition of Runna, a UK-based fitness app specializing in AI-powered training plans and coaching. While Strava’s plans for the product remain somewhat unclear, there’s obvious overlap between the two platforms’ features. It’ll be interesting to track whether their tools are eventually integrated or continue to operate separately.