How to Learn From Your Previous Spotify Releases

You learn from your previous Spotify releases by pulling the same handful of numbers for each of your last three or four releases, putting them side by side in one table, and picking one thing to change next time. It takes about an hour the first time, and it tells you more about your own music than any general advice can. Here's what to pull, what to ask it, and what the common patterns mean.

Pull the same numbers for every release

Everything below is in Spotify for Artists. Open each of your last three or four releases and write down, for the first 28 days after release:

• Streams and listeners. A stream counts once someone plays at least 30 seconds, according to Spotify's guide to how streams are counted.
• Saves. Someone liked the track enough to keep it. This is the number that matters most.
• Streams per listener. How many times each listener came back to it.
• Followers gained in the same window.
• Where the streams came from. Under Segmentation, then Source of Streams, Spotify splits them into active sources, like your profile and listeners' own libraries, and programmed ones, like editorial, algorithmic, and other listeners' playlists. Spotify's guide to source of streams explains each category.
• The playlists you were on. Music, then Playlists, lists them with how many listeners each one sent.

Then add the things only you know: what you spent and on what, which playlists added you and how many followers each had, when you started promoting, and anything that was different about that release.

Put them in one table

One row per release, one column per number, oldest release at the top. A single release tells you very little on its own. The comparison down each column is where the answers are.

Say your table looks like this for three singles:

• Release 1: 4,000 streams, 90 saves, 20 new followers, $150 spent. $1.67 per save.
• Release 2: 9,000 streams, 110 saves, 25 new followers, $300 spent. $2.73 per save.
• Release 3: 6,000 streams, 240 saves, 70 new followers, $150 spent. 63 cents per save.

Release 2 looks like the winner on streams. Release 3 is the one that worked: fewer streams, but more than twice the saves and followers, at a quarter of the cost per save. Whatever was different about release 3 is what you repeat.

Five questions to ask your table

1. Which release did best on saves, and what was different about it?

Look at the track before you look at the money: the tempo, the length, how fast it gets going, and how clearly it sits in one genre. Then look at the timing and which playlists it reached. Tracks that earn saves usually share something, and it's often something you can do again on purpose.

2. Where did your listeners come from?

If most streams came from active sources, your existing fans are carrying the release, and your promotion reached few new people. If most came from programmed sources but saves stayed low, you reached plenty of new listeners who didn't connect with the track, which usually points at playlists that were the wrong fit.

Spotify also sorts your audience into segments, including super listeners who played your music on purpose 15 or more times in 28 days, in its guide to audience segments. A release that grows that group is building fans.

3. Which playlists delivered, and how big were they?

Usually smaller ones than you'd guess. Start with every playlist that added you, and its follower count. The range those counts fall in is your realistic target for next time. The words those playlists use to describe themselves are also how curators hear your genre, so use them when you set the genres for your next campaign.

Then list the playlists that actually sent listeners. A playlist only shows up in Music, then Playlists, once it has sent at least 3 listeners, according to Spotify's guide to the playlists you're added to. A small playlist with engaged listeners often beats a big one that sent almost nobody. How to tell if a playlist has real listeners shows you what to check.

4. Did the timing make a difference?

Usually, yes. Compare the releases you promoted from release week with the ones you got to late. The gap is often big, and closing it costs nothing but planning.

5. What did each release cost per save?

Divide what you spent by the saves. A save means someone chose to keep the track, so cost per save tells you whether your money reached people who liked it. How to measure Spotify playlist promotion results goes deeper on each number.

What the patterns usually mean

• Streams flat, saves rising. Your music is connecting better, and reach is the limit. Do more of the reach work.
• Streams rising, saves flat. Your promotion is reaching the wrong listeners. Narrow the playlists you target.
• Followers flat across every release. Listeners arrive and leave. Look at your profile and at what happens after release week.
• One release far ahead of the rest. Work out what was different and do it again deliberately.
• Everything flat. The track is the constraint, and more promotion won't move it. Is your track ready for playlist promotion is the place to start.

Turn it into a brief for your next release

By now you should have written down:

• Your genre in three words, in the words the playlists that added you use.
• The follower range of playlists that realistically add you.
• The playlists that sent real listeners.
• One change to make on the next release.

Stick to one change. Change five things and the next row of the table can't tell you which one worked.

That brief is also exactly what a Soundmango campaign asks for. You start a campaign by setting your genres and the minimum follower count from your range, see one price, and pay once. Your track only goes to curators whose playlists fit, and it keeps going to the next fitting curator until one takes on your campaign, so approved tracks have an estimated 99% chance of landing on at least one playlist.

Keep the record from now on

The review is hard the first time because nobody collected the numbers when they were fresh. From your next release on, add a row six weeks after release day. It takes ten minutes, and after four or five rows you'll know which kind of playlist, which size, and which timing works for your music.

Frequently asked questions

How far back should I look?

Three or four releases, or roughly the last two years. Older releases describe a different version of your music and your audience, so weight the recent ones most.

Is it worth doing if I've only released twice?

Yes. Two releases already show you the genre words and the playlist size that said yes to you. Start the table now, and every release after this one adds a row.

What if none of my releases got onto any playlists?

Then start with the track and the targeting. Compare your track with a released track in your genre at the same volume, and narrow your genre to three words. Why your playlist pitches aren't getting placements covers the most common reasons a track gets passed over.

Which number matters most?

Saves in the first 28 days. It's the closest thing you have to listeners telling you the track worked.

What if my releases are years apart?

Compare them anyway, and count the gaps as part of what you learn. Irregular releases start colder each time, because fewer listeners remember you.

Make your next release the one you learned from

Your own numbers already say which playlists and which listeners suit your music. Put that brief into a Soundmango campaign, pay once, and your track goes to curators in your genre until one takes it on, and if no curator does, you get your money back. Start a campaign for your next release.