FAQs

Topics we frequently receieve questions on are answered here.

Why can't I authorize my Strava account?

Occasionally we see errors in our logs indicating a user can't get past the initial authorization step. This happens if your first or last name in your Strava profile is longer than 60 characters.

The experience for Mobile users in this situation is they never receive a welcome confirmation email or have any activities show up in their feed.


What is the best setup for runners?

While we support heartrate and pace related metrics and achievements through our Strava relationship, authorizing us to recieve your raw workout files from Garmin Connect significantly enhances the value we can provide.

  1. Access to Garmin Running Dyanmics and Stryd Metrics comes from Garmin Connect.
  2. Strava began to remove running power from activities pushed from Garmin Connect in July of 2017.
  3. The running specific variables (e.g. Ground Time) have a more material impact on your pace than the 12 standard variables we track from our base factors sourced from Strava.
  4. If you don't want to analyze your running specific variables there are a few workarounds to the Strava Run Power issue. You can pair Stryd via Ant+ and record the activities as Ride and then change them to Run in Strava after uploading, or you can upload the raw fit files to Strava manually.
  5. We've seen that Polar+Stryd users don't loose their run power data within Strava. Issue here is Polar doesn't record the running specific variables (other than power).
  6. Last but not least, if you use Humon or Moxy to measure muscle oxygen we gain access to this data from Garmin Connect and we'll include it on your Running Pace factors automatically if it's present.

How do I get the best results from the Bike Classifiers?

Best results are obtained if you authorize strive.ai to receive your raw activity files via Garmin Connect.

This allows us use the actual sensors on each bike to train our predictors. The standard predictor uses ride telemetry (avg speed, elevation, duration, etc) to predict. This one works for many, but not all use cases.

  1. The standard version works best if you have multiple bikes you ride with unique telemetry (e.g. MTB vs. Road Bike). If you have two road bikes or a TT bike and a road bike it may struggle to distinguish between the two.
  2. It's best to have 1 or 2 unique sensors (e.g. Cadence, Speed or Power) on each bike.
  3. Make sure you only use one label in Strava for each physical bike (e.g. Road Bike not (Road Bike and Virtual Road Bike).
  4. Ensure the labels (bike assigned to each Strava activity) are correct for the past 8 weeks in Strava prior to signing up with strive.ai and/or for the first 5 rides on each bike after authorizing Garmin Connect.
  5. Most importantly reach out to us at info@strive.ai if you can't get it to work and we'll help work through any problems.
  6. We may reach out to you if you thumbs down a handful of predictions to help resolve also!
  7. Pro tip: We train these predictors every two weeks using all your history for the standard classifier and the most recent 5 rides on each bike for the Garmin classifier.