Spotify’s is giving you the keys to its recommendation algorithm with US launch of ‘Taste Profile’
TechCrunch Sarah Perez
Spotify is letting U.S. Premium users edit their taste profile in plain English. That means the app’s recommendations can be steered, not just passively accepted.
Based on reporting by TechCrunch, Sarah Perez — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Spotify is handing Premium listeners in the U.S. a new way to poke at the machinery behind its recommendations. The feature, called Taste Profile, started rolling out on Wednesday and shows users how Spotify thinks about their listening habits across music and other audio, then lets them ask for changes with natural-language prompts.
This is more than a settings tweak. Taste Profile feeds the same recommendation engine behind the Home feed, Discover Weekly, Made for You picks, and Spotify Wrapped, so changing it can shift a lot of what shows up next. If the app has been stuck recommending more of the same, users can ask for more variety, or push harder toward or away from a genre, artist, or general vibe.
Spotify first showed the feature in March at SXSW in Austin, with co-CEO Gustav Söderström attached to the reveal. The company then tested it in beta in New Zealand. Now it’s arriving in the U.S., which Spotify describes as its biggest market for traffic and revenue, and the rollout is aimed at listeners 18 and older.
There’s a practical use case here that goes beyond music nerd tinkering. Spotify says listeners can use it to shake off sleep sounds, clean up the effect of music a teenager plays on a parent’s account, or steer podcast and audiobook recommendations too. After a few hours, the Home page should reflect the edits. The feature is still in beta, and Spotify hasn’t said when it will be pushed to everyone.
My take — AI-written commentary, not fact-checked reporting
Spotify is doing the sensible thing here: letting users see and shape the machine instead of pretending the machine is magic. That’s the rare platform move that actually respects taste, rather than turning every habit into a forever recommendation. More apps should stop acting like users are trapped inside their own bad data.
Read more about this at: TechCrunch
