Spotify is prototyping Tastebuds, a social feature that allows users to interact with and discover music by searching the people they follow
Context & Ripple Effects
Tastebuds is not an isolated experiment but the next move in a multi-year push to make Spotify's social graph do work that algorithms currently do. Mid-2019 brought reports of Social Listening, a shared queue friends could fill together, which later shipped as Group Session for Premium users in May 2020. Even earlier, Bumble let dating profiles surface music taste via Spotify accounts, evidence that listening history was already functioning as identity.
The throughline matters because discovery is Spotify's most valuable real estate: whatever surface decides what plays next captures the listener. Tastebuds points that surface at the people a user already follows rather than at editorial playlists or recommendation engines.
First-order effects
- Users gain a people-based discovery channel — searching followed accounts to find music — putting friend curation in direct competition with Spotify's own algorithmic and editorial playlists for the same listening time.
- Spotify gets another retention lever built on its existing follower graph, at near-zero content cost since it reuses data the platform already has.
Second-order effects
- If people-based discovery proves sticky, curators inside the network — friends, tastemakers, playlist builders — become a recommendation layer Spotify can promote or monetize, shifting internal power away from purely algorithmic surfaces.
- Rival streaming services without an active in-app social graph face pressure to either build one or partner for it, repeating the pattern Bumble chose when it connected profiles to Spotify rather than building music features itself.
Third-order effects
- The cadence here — Social Listening reported in 2019, Group Session rolled out in 2020, Community tested in 2022, Tastebuds prototyped alongside them — suggests Spotify treats social features as standing R&D, shipping whichever experiments stick, which would make the social graph a durable moat on top of licensed music any competitor can also license.
- If human-curated discovery keeps gaining share of listening sessions, the industry's attention economics shift from whoever owns the catalog toward whoever owns the connections between listeners — an access-layer contest rather than a content one.
The trend: Music streaming is shifting discovery from algorithmic playlists toward human social graphs, with Spotify running a rolling series of social experiments to find which ones stick.