Title: Running on the Beat (music + fitness)
Summary: Starting with a library of songs that a user already likes, create playlists whose songs have a BPM (beat per minute) that matches a person’s running pace. For example, if a user typically runs at 6mph (10 min mile), generate a playlist with songs at roughly 162 bpm
Data sources: Strava, Spotify
General Description of Idea: Analyze user’s Strava data to figure out how fast a user runs (this is given in the “activity” log file). Then, analyze user’s Spotify data to generate a large library of all songs the user has streamed (“all” can just be top 1000). Finally, generate playlists full of songs whose BPM matches the average speed at which the user runs (this will require looking up the BPM of each song, but this is relatively easy).
Purpose: Running “on the beat” somewhat magically tends to give a person more energy and motivation. The idea of music+cardio is extremely well-established and goes back many decades (Jazzercise, Aerobics, Zumba, and more recently Soul Cycle, Peleton). However, mainstream apps like Strava do nothing to help a user play music that matches his or her pace. We can fix that problem by generating appropriate playlists for users, using music the user already loves. At first, these can be extremely simple, single speed / single beat playlists. Later, we can take the min, avg, and max speed at which a user runs and build a playlist with all songs at all three levels of speed (start with min, then move to average speed, then finally to max speed, and then back down again - there’s your workout).
(Note: I have intentionally kept this template very light on details because my goal is to get novel ideas from lots of people, not detailed explanations of how to implement the idea - we can work on that later. You don’t need to know exactly how an idea will work in order to propose it).
I’d love to hear what ideas you have. What data integrations do you wish existed? Use this template if you find it helpful, or feel free to express your idea however you like. At a minimum, though, it would be nice to include the data sources you envision using to power your idea.