Recommendation System Users Can Adjust Themselves

Music streaming platform Spotify has unveiled a new tool that allows users to view and edit their own taste profiles directly. Co-CEO Gustav Söderström announced the feature last Friday at SXSW in Austin, stating that a beta version will be released within weeks to Premium subscribers in New Zealand.

The core of this tool is recommendation adjustment through natural language commands. Users can modify their taste profiles built by the algorithm with simple sentences like "recommend more 90s hip-hop" or "reduce podcast recommendations." These changes are immediately reflected in Discover Weekly, home feed recommendations, and year-end summary feature Wrapped results.

Enhanced Transparency in Recommendation Algorithms

This feature is interpreted as Spotify's attempt to increase transparency in its recommendation system. Previously, users could only perform limited manipulation by excluding specific songs or playlists from their profiles. However, the new tool visualizes a comprehensive taste profile integrating listening data across music, podcasts, and audiobooks, allowing users to edit it directly.

This is a response to long-standing complaints that streaming platform recommendation algorithms operate like "black boxes." It grants control to users who couldn't understand why algorithms recommended certain songs or how their tastes were categorized.

Intensifying Personalization Competition in Streaming Market

Spotify's move demonstrates that personalization technology competition among streaming platforms is entering a new phase. Competitors like Apple Music and YouTube Music have each strengthened their recommendation algorithms, but a feature allowing users to directly edit taste profiles is an industry first.

Spotify plans to expand this feature globally beyond New Zealand based on beta test results. However, whether it will be offered to free users or maintained as a Premium subscriber-exclusive feature has not yet been determined.