Short answer: The Spotify algorithm recommends songs mainly from each listener’s “taste profile”: what they play, skip, search for and save, plus signals such as the songs people add to playlists and what similar listeners enjoy. It powers personalised playlists like Discover Weekly and Release Radar, Radio, Autoplay and Mixes. Spotify also says commercial factors, including its paid Discovery Mode tool, can influence recommendations, while editorial playlists are chosen by people.
There is no single “Spotify algorithm” switch to flip. There are several recommendation surfaces, each fed by similar signals but built for a different job. Knowing which is which helps you put effort where it actually changes something.
Which Spotify playlists and features are algorithmic?
| Surface | How it works | What you can influence |
|---|---|---|
| Release Radar | New releases each Friday from artists a listener follows, listens to or is predicted to like | Followers, pitching at least 7 days before release |
| Discover Weekly | A personalised discovery playlist that updates every Monday | Indirectly, through genuine listening, saves and playlist adds |
| Radio, Autoplay and Mixes | Continuous, personalised listening based on a song, artist, mood or the listener’s habits | Listening signals; Discovery Mode applies here |
| Editorial playlists | Curated by Spotify’s editors (Spotify logo in the byline) | Pitching one unreleased song; no guarantee |
| Listener playlists | Made by users themselves | Fan outreach and your own playlists |
Spotify labels its personalised playlists “Made for [username]”. For some of them, editors choose the pool of songs and the algorithm decides which of those each person hears.
What signals does the Spotify algorithm use?
Spotify’s own explanation of its recommendations lists four broad groups of input:
- The listener’s taste profile: built from searching, listening, skipping and saving to Your Library. Spotify calls this the most important input.
- Information people share: general location, device, language, age and who they follow.
- Trends: how other users behave in aggregate. If many people engage with something, it may be recommended to similar listeners.
- Content characteristics: such as genre and release date.
Spotify’s artist resources add that its personalised playlists look at what and when people listen, which songs they add to their playlists and the habits of listeners with similar taste. That is why saves and playlist adds matter: they are among the clearest signs that a listener wants more of a song. Spotify does not publish weightings, exact thresholds or a “save rate” that triggers anything, so treat any precise formula you see online as guesswork.
Do skips hurt a song?
Spotify confirms that skipping is part of a listener’s taste profile, so it shapes what that person is shown later. It does not publish how skips are weighted. A skip also means the play may not reach 30 seconds, the point at which Spotify counts a stream (see what counts as a stream on Spotify).
How do Release Radar and Discover Weekly choose songs?
Release Radar
Release Radar refreshes every Friday. Your followers receive songs from your new releases. If you pitch a song through Spotify for Artists at least seven days before release, that is the song your followers get; otherwise Spotify picks one. Other rules from Spotify’s help centre:
- Listeners get at most one song per artist per week.
- A song can stay for up to four weeks if the listener has not played it yet.
- Remixes can be included, but acoustic, karaoke and live versions cannot.
- Re-releases of songs already on Spotify and Various Artists releases are excluded.
Because followers are the core audience here, anything that adds genuine followers before release helps. Pre-save links that also follow you on Spotify are one way to do that; see how Spotify pre-saves work.
Discover Weekly
Discover Weekly updates every Monday with music a listener has not necessarily heard yet. In 2025 Spotify added genre controls at the top of the playlist for Premium listeners, letting them steer it towards up to five personalised genre options. You cannot pitch to Discover Weekly. Your songs reach it when listening data suggests people with similar taste will enjoy them.
What is Spotify Discovery Mode?
Discovery Mode is a Spotify for Artists tool, still in beta, that lets eligible teams flag songs they want recommended more often. Spotify says it raises the likelihood of those songs being recommended but does not guarantee it. Key points:
- Where it applies: Spotify Radio, Autoplay and Spotify Mixes (Daily, Artist, Mood, Decade and Genre). It is not active in editorial playlists.
- Cost: no upfront budget, but a 30% commission on recording royalties from streams of selected songs in those contexts. Other streams are commission-free.
- Eligibility: among other rules, at least 25,000 monthly listeners and three eligible songs; songs must have been out for at least 30 days, met Spotify’s monetisation criteria the previous month and come through a participating distributor.
- Timing: campaigns are set up during a monthly window and run from the 1st of the following month.
Spotify states openly that commercial considerations, such as the cost of content or whether it can be monetised, may influence its recommendations. Discovery Mode is the most visible example.
How can you work with the Spotify algorithm?
- Pitch every release. One unreleased song at a time, at least seven days ahead. It may not land an editorial spot, but it decides which song reaches followers’ Release Radar.
- Grow followers, not just plays. Followers are the audience Release Radar is built on.
- Send listeners who will stay. Promotion aimed at people who like your genre produces saves and repeat plays; untargeted traffic that skips sends weaker signals.
- Keep releasing. Regular releases give Release Radar something new to deliver.
- Read Source of streams. Spotify for Artists splits your streams into active sources (your profile, listeners’ own playlists and Liked Songs) and programmed sources (editorial, algorithmic playlists and Radio), so you can see which surfaces are working.
For promotion tactics beyond the algorithm, our complete guide to getting more Spotify streams goes further. If you are comparing growth options, our page on which Spotify service fits which goal explains the differences honestly; artificial streams that Spotify detects are removed and earn nothing.
Frequently asked questions
How long does it take for the Spotify algorithm to pick up a song?
Spotify does not give a timeline. Release Radar can include a new song in its first Friday if it is delivered at least seven days before release, while Discover Weekly and Radio depend on listening data building up over time.
Can listeners turn off algorithmic recommendations on Spotify?
Partly. Listeners can switch off Autoplay and Smart Shuffle in settings, hide a song in a playlist, and Premium users can snooze a track from recommendations for 30 days.
Does the Spotify algorithm favour bigger artists?
Spotify does not say so directly, but its recommendations use aggregate trends, and tools such as Discovery Mode require at least 25,000 monthly listeners. Smaller artists still reach Release Radar through their followers and can pitch to editors.
Is Discovery Mode the same as paying for a playlist?
No. It does not place songs on editorial playlists, and it guarantees nothing. It is a recommendation signal for Radio, Autoplay and Mixes, paid for through a commission on royalties from those streams.
Sources: Spotify: Understanding recommendations, Spotify for Artists: Getting music on Release Radar, Spotify for Artists: Using Discovery Mode, Spotify Newsroom: Discover Weekly turns 10