Spotify recommendation signals come down to four inputs, named on one public page. Your taste profile leads. Then the details Spotify holds about each listener, what other listeners do, and the song’s own characteristics. No save rate percentage appears there. Two of the four show up in reports you already own.
In this guide you’ll learn.
- The four inputs. What Spotify’s own page names as a recommendation signal.
- The taste profile. Which listener actions feed it, including the one that subtracts.
- The reports. Where each signal shows up in Spotify for Artists.
- The paid signal. What Discovery Mode adds and what it costs.
- The limits. Which numbers Spotify has never published.
- The mistakes. Where the advice on page one goes wrong.
On this page
- What Are Spotify Recommendation Signals?
- Searching, Playing, Skipping and Saving Build the Taste Profile
- Language, Age, Location and Follows Decide Who Gets the Recommendation
- What Other Listeners Do With a Search Result Counts as a Signal
- Genre and Release Date Are Signals Before Anyone Presses Play
- A Premium Listener Can Snooze Your Song for 30 Days
- Discovery Mode Adds a Signal and Spotify Charges a Commission on It
- Do Playlist Adds Change What Spotify Recommends?
- Saves and Follows Show in Your Reports and Trends Never Will
- Spotify Publishes No Save Rate Threshold for Recommendations
- Common Mistakes
- FAQ
- Conclusion
- Next Step
What Are Spotify Recommendation Signals?
They’re the inputs Spotify’s algorithms read when they choose and order what a listener sees. That covers Search, Home and personalized playlists. Spotify’s safety and privacy centre names four of them in an article updated in March 2026. Those four are your taste profile, the details Spotify learns about you, trends across other listeners, and the characteristics of the content itself.
One line under that list does more damage to the advice on page one than anything else. Spotify writes that the importance of each input shifts over time and by listener. So a single threshold that works for every song cannot exist.
Searching, Playing, Skipping and Saving Build the Taste Profile
Spotify names four listener actions under the taste profile. Searching, listening, skipping and saving to Your Library. It calls the taste profile the most significant input for the overall experience. Three of those actions help you. The fourth is a skip.
Skipping is the one almost nobody writing about this mentions, and it’s the one you can do least about after release. Spotify for Artists reports no skip rate at all, so reading it directly is out. What you can read is the ratio between streams and listeners on a song. It moves in the same direction.
A save has a specific definition worth knowing, since the word gets used loosely. Spotify counts a save when a listener taps the plus button and the song lands in their library. A play on a playlist is something else. So is a follow of your artist profile.
All three sit in different rows of your reports, which is why reading a save against the add date works at all.
Language, Age, Location and Follows Decide Who Gets the Recommendation
The second group covers what a listener told Spotify rather than what they played. Spotify’s page names general location, device, language, age and who the listener follows. Pick German as your language and German language recommendations follow. Follow an artist and their new music arrives.
Only one entry in that group is yours to influence. It’s the follow. Everything else describes the listener. That gap is why the follower count on your profile behaves differently from your monthly listener count. The two numbers answer different questions about your audience.
Location is worth one caution. Spotify describes it as general and non precise. You’re looking at a country and a broad region rather than a city block.
What Other Listeners Do With a Search Result Counts as a Signal
Spotify’s third input is trends. It defines those as the general tastes and behaviours of listeners as a group. The example on the page is a search result. When many listeners interact with one result, Spotify becomes likelier to serve that result to other people searching for similar things.
This is the input that runs on volume. It’s also the one an artist has the least direct hold on. You can’t make a thousand strangers search the same phrase. What you can do is make sure the phrase leads somewhere, which is a metadata job rather than a promotion job.
Artificial streaming aims straight at this input. That’s why Spotify’s policy page says flagged streams carry no positive weight in recommendations. Bought activity puts fake data into the trend count. Spotify discounts the streams and the penalty still lands on the release.
Genre and Release Date Are Signals Before Anyone Presses Play
The fourth input is the song itself. Spotify names genre, release date and category. It says this is how it works out which songs share characteristics and might suit similar listeners.
Both of those fields arrive through your distributor. The release date sets your position in Release Radar, where the order runs on release date and predicted taste. Your genre tag decides which songs yours gets grouped with. All of that happens before a single listener has heard it.
So the tag is a decision rather than a formality. A broad tag drops your song into a much larger pool. Pick a precise subgenre and it sits beside twenty artists a curator in that subgenre already plays. Same reason the tag is worth checking before you screen playlists.
A Premium Listener Can Snooze Your Song for 30 Days
Spotify publishes the subtracting controls in more detail than the adding ones. Most artists don’t read that part.
A listener can exclude a playlist or a track from their taste profile. The excluded track then carries less weight in what Spotify serves them. A thumbs down brings fewer recommendations like it.
Hiding goes further. Tap hide on a track and it disappears from that playlist on every device the listener uses, so it won’t play there again. A Premium listener can then snooze it. Spotify puts that at 30 days across the whole service.
Skips, hides and snoozes are the reverse of a save, and Spotify documents all three. A song placed in front of the wrong audience collects them, which is the argument against a wide send that stream counts never make.
Matching the list to the song is what keeps those controls out of play. On PlaylistFeed you filter verified curators by genre and subgenre before anything goes out.
Discovery Mode Adds a Signal and Spotify Charges a Commission on It
Spotify’s recommendations page is direct about commercial influence. It says commercial considerations can affect recommendations, and it names Discovery Mode as the example. Turning it on for a song adds that song’s priority to the algorithms behind personalized listening sessions.
Three limits sit in the same paragraph. Spotify charges a commission on streams in the areas where Discovery Mode runs. Discovery Mode stays out of editorial playlists. The signal raises the chance of a recommendation without promising one, because Spotify only serves songs it expects listeners to enjoy.
Spotify says it notes when a listener isn’t engaging with a song, Discovery Mode songs included, and reads that into later recommendations. So a song people skip stays a song people skip, commission or not.
Do Playlist Adds Change What Spotify Recommends?
Yes, through the listeners the add reaches rather than through the add itself. Spotify’s page on playlist types describes how algorithms build personalized playlists. They read what a listener plays and when, which songs land in that listener’s own playlists, and the habits of similar listeners.
An independent curator adding your song puts it in front of listeners. Those listeners take one of the taste profile actions or skip past it.
Your Source of streams report separates the two sides. A curator playlist sits on the programmed side while the run is happening. The listeners who kept the song turn up afterwards on the active side, in the row covering their own library and playlists.
Finding curators who run lists your song fits is the part that happens first. On PlaylistFeed you search verified curators by genre and send the pitch from the same screen. Making an account and searching the verified playlists is free. Pitching is on the paid plan.
The signals above only start once somebody presses play. A genre matched add is what puts the song in front of somebody who might.
Saves and Follows Show in Your Reports and Trends Never Will
Two of the four inputs have a report behind them. The other two describe the listener and the crowd. No artist facing screen reports either one.
| Signal | Where you see it in Spotify for Artists | What it will not tell you |
|---|---|---|
| Taste profile actions | Saves, and the library row of Source of streams | Which listener skipped, and when |
| Information about the listener | Follower count, and the country breakdown in Audience | Language, age or device of any listener |
| Trends | No report | Anything about search behaviour |
| Content characteristics | Genre and release date on your release page | How Spotify groups your song against others |
The Playlists tab is the closest thing to a signal report you have. It ranks the top 100 playlists carrying your music by the listeners each one brought over the last 12 months. Three of your listeners have to be on a playlist before it appears at all. So a small add can be real and it’s still invisible to you.
Spotify Publishes No Save Rate Threshold for Recommendations
Search this query and a save rate figure comes back as the trigger, usually around twenty percent, with no source under it. Skip rate and completion rate turn up beside it, described as metrics you can track. Spotify for Artists shows neither of those as a number.
Four things stay unpublished. There’s no weighting across the four inputs. No save rate, skip rate or completion rate threshold. And no stream count that moves a song into Discover Weekly.
Spotify’s editors have said separately that follower and monthly listener counts stay out of editorial decisions. That’s a human system rather than the algorithmic one. It still gets quoted as though the two were the same.
So the honest read is a comparison against yourself. Take the save count and the streams to listeners ratio on your last release. Then run the same two on this one at day 28.
That comparison depends on a few hundred monthly listeners sitting behind it. Under that, a single playlist add swings the ratio far enough to hide the rest.
Common Mistakes
- Chasing a percentage. Working toward a save rate figure Spotify has never published as a threshold.
- Reading a skip rate. Spotify for Artists reports no skip metric, so a figure like that came from somewhere else.
- Buying activity. Flagged streams carry no positive weight in recommendations and the penalty lands on your release.
- Tagging broad. A general genre groups your song with a pool too large for a curator in your subgenre to search.
- Sending wide. A song served to the wrong audience collects skips and hides, which are inputs too.
- Treating editorial as algorithmic. The pitch form reaches human editors, and that route closes once the song is live.
FAQ
Does Spotify rank one signal above the others?
Its page calls the taste profile the most significant input for building the overall experience. Spotify adds that the importance of every input varies over time and by listener. There isn’t a published weighting beyond that.
Is there a save number that gets a song recommended?
Spotify publishes no such number. The figures circulating on this query come from third party campaign write ups, and none of them names a source.
Do skips work against a song?
Skipping sits inside the taste profile alongside listening and saving, so Spotify reads it as an action. It describes hides and snoozes the same way, as feedback that reduces similar recommendations.
Can I see which listeners skipped my track?
No report in Spotify for Artists names a person or their skips. The streams to listeners ratio on a song moves in the same direction and is the closest proxy available.
Does Discovery Mode guarantee a recommendation?
Spotify states that the signal increases the likelihood of a recommendation without promising one. It’s still reading listener engagement on those songs afterwards.
Do editorial playlists use the same signals?
Human editors put editorial playlists together using data, listening and cultural knowledge. Discovery Mode stays out of them, and Spotify’s editors say follower counts sit outside their decisions.
How long before a new release affects recommendations?
Streams start counting at 30 seconds from the day the song is live. Discovery Mode needs a song live for at least 30 days first.
Conclusion
Spotify names four inputs on one public page and puts no threshold on any of them. The Spotify recommendation signals an artist can act on are the taste profile actions and the genre and release date you deliver. Both need a listener who’s already found the track. Trends and listener information describe an audience you can reach rather than a dial you can turn.
Open Source of streams and the Playlists tab after your next release. Read them against the same two reports from your last one. That comparison is the closest thing to a signal report Spotify gives an artist. It beats any percentage quoted on a blog with no source under it.
Next Step
Every signal above starts with a listener hearing the song, and most new listeners never will. Playlists are where a stranger hears it. A curator decides what goes on each one.
On PlaylistFeed you search verified curators in your genre, send the pitch before release day and track who replies. Searching is free, and every playlist is verified before it goes up.
Want to see which curators in your genre are adding songs right now?
Find curators in your genre and start with the lists already playing music like yours.
