Spotify Hands U.S. Premium Users Control Over Recommendation Algorithm With 'Taste Profile'

Feature unveiled at SXSW in March arrives in the company’s largest market after a beta test in New Zealand

By LineZotpaper
Published
Read Time2 min
Sources2 outlets
Spotify is rolling out a new feature called 'Taste Profile' to Premium subscribers in the United States that lets users see how the platform understands their audio preferences and adjust recommendations using natural language commands. The AI-powered tool, first previewed at SXSW in March by co-CEO Gustav Söderström and tested in New Zealand, effectively gives listeners direct control over the algorithm that powers Discover Weekly, Made for You playlists, and Spotify Wrapped.

The music streaming giant on Wednesday began rolling out Taste Profile to U.S. Premium subscribers aged 18 and up. The feature addresses a common complaint: Spotify’s recommendation engine, while widely praised for its accuracy, often fails to keep pace with evolving tastes, repeatedly suggesting songs too similar to what a user has already listened to.

With Taste Profile, users can view a breakdown of how Spotify categorises their preferences across music and other audio content. They can then make adjustments using natural language requests—telling the platform they are currently more interested in a specific genre or want fewer podcast recommendations, for example. The platform then updates its recommendations accordingly across the Home feed and key personalised playlists.

Spotify’s recommendation algorithm has long been considered a competitive advantage, but the closed nature of that system has meant listeners have had little recourse when suggestions feel stale. By opening the engine up to direct input, Spotify is attempting to give users more agency without sacrificing the convenience of automated discovery.

The rollout to the U.S., Spotify’s largest market by traffic and revenue, marks a significant expansion of the feature. The beta in New Zealand provided early data on user behaviour, and the U.S. launch will test whether the concept scales to a broader and more diverse audience.

§

Analysis

Why This Matters

  • For the first time, listeners can directly influence the algorithm that decides what audio appears in their feeds, potentially breaking users out of recommendation loops.
  • The feature could reshape the music discovery landscape, giving artists and podcasters a new way to be found if listeners explicitly opt into genres or moods.
  • If successful, Taste Profile may set a precedent for greater user transparency and control across other algorithmic platforms.

Background

Spotify’s recommendation system is a core part of its product, driving playlists like Discover Weekly that have become iconic for the service. However, the system has drawn criticism for creating "filter bubbles" where listeners hear only music very similar to their established tastes. Taste Profile is Spotify’s attempt to address that by letting users intentionally steer their recommendations toward new directions without having to manually search or break their listening habits.

Key Perspectives

Users: Gain visibility into how their tastes are categorised and the ability to fine-tune recommendations. However, the feature adds complexity to what was previously a passive experience. Spotify: Benefits from increased engagement and satisfaction, plus richer data about user intentions. The company also strengthens its AI-driven personalisation narrative. Artists and Labels: Could see new opportunities to reach listeners who explicitly express interest in a genre or style, but may also worry about being deprioritised if users narrow their preferences.

What to Watch

  • How many U.S. Premium users actually engage with the feature in the first month.
  • Whether Spotify expands Taste Profile to other markets or to its free ad-supported tier.
  • Feedback on whether the natural language commands actually correct stale recommendations or introduce new frustrations.

Sources

Zotpaper

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.