How Discover Weekly picks tracks (for the artist side)

Discover Weekly is Spotify's most powerful algorithmic playlist for independent artists. The exact model is proprietary, but the signals it responds to are consistent and observable.

Collaborative filtering, not just audio

Discover Weekly leans heavily on what other listeners with similar taste are saving and playing. A track saved by 300 listeners who also save music similar to a target user is much more likely to appear on that user's list than a track with 3,000 plays and no saves.

Save rate over play count

Save-to-listen ratio is one of the clearer predictors. Tracks under three minutes with high completion and save rates punch above their weight; longer tracks with high skip rates rarely enter algorithmic rotation.

Freshness window

Tracks released in the last 30 days get an implicit boost in algorithmic consideration. Missing that window with slow promotion is one of the most common missed opportunities for indie releases.

Cross-track behaviour

Listeners who go from your track to another song on your profile — rather than skipping out of Spotify entirely — send a strong signal. A well-ordered artist profile (top tracks visible, related singles nearby) helps this happen without extra promotion.

For sustained streaming numbers, see the Spotify promotion service and the cross-platform marketing pillar.