Discover Weekly Has No Submit Button. Here's What Actually Gets You On It.

There's no submit button for Discover Weekly. No pitch form, no inbox, no curator to charm into a yes. I wish I had something gentler to open with, but that's the honest starting point, and anyone promising guaranteed placement is selling you something that doesn't exist in that form. (We've laid out who's actually selling what in our honest rundown of Spotify promotion services, if you want the receipts.)

Here's what most "how to get on Discover Weekly" posts skip past too fast, though: the algorithm isn't a sealed box you have zero influence over. It's downstream of things you actually do. Understand the chain, work the links you control, and stop paying for the ones you can't.

There's no pitch, but there is a chain

Discover Weekly runs on collaborative filtering. In plain terms: Spotify groups listeners into taste clusters based on overlapping habits, and when two listeners share enough of that overlap, a song one of them plays regularly becomes a candidate for the other person's Discover Weekly. That's the whole trick. The algorithm isn't listening to your song and judging whether it's good. It's watching whether people who already like artists similar to you are starting to engage with your track the way they engage with those artists.

Which means the real question isn't "how do I get on Discover Weekly." It's "how do I get my track in front of the right taste cluster long enough to generate the engagement that proves the match." Those signals have to come from somewhere, and editorial placements are a big part of it, since a spot on an editorial playlist exposes your song to thousands of listeners and generates the engagement data that feeds both Release Radar and Discover Weekly. Independent genre playlists do the same thing at a smaller scale. That's exactly why playlist placement work still matters even though nobody can pitch Discover Weekly directly.

There's a second layer under collaborative filtering too. Discover Weekly also runs content-based audio analysis of a track's actual sonic features, plus natural language processing of what's written about an artist online. That matters for brand-new artists specifically: a new artist with zero prior streams can still land in someone's Discover Weekly largely through those content and NLP layers, which don't need existing listening history the way collaborative filtering does. So day one isn't a dead zone. It's just a different mechanism doing the work until you've built enough listening history for collaborative filtering to take over.

Why your own fans go first

Here's a pattern I've watched play out again and again with artists we work with: Discover Weekly doesn't trust strangers' opinions until it trusts yours first. Think of it as a quiet QA step before the wider release. The system watches how your actual fans react to a new song before it starts routing that song to strangers in a similar taste cluster. If your core listeners stream a release once and vanish, that's a signal too, just not the one you wanted. Build the base. Let the algorithm extend outward from there, not the other way around.

Discover Weekly vs. the other playlists you'll hear about

People lump all of Spotify's playlist systems together, but they run on different fuel. Knowing the difference stops you from chasing the wrong metric.

Playlist	What it actually is	Primary signal	Can you pitch it?
Discover Weekly	30 songs a listener has never heard, refreshed every Monday	Collaborative filtering + content/NLP analysis	No
Release Radar	New music from artists a listener already follows	Follow status + engagement	No
Editorial playlists (RapCaviar, Fresh Finds, etc.)	Human-curated genre and mood playlists	Editor taste, submitted via Spotify for Artists	Yes, via pitch tool
Independent/curator playlists	User-made playlists, often genre-specific	Curator taste, direct outreach	Yes, directly

Release Radar surfaces new music from artists someone already follows. Discover Weekly introduces songs from artists they've never heard. Radio just extends a session with similar tracks once momentum is already going. Editorial and independent playlists are the only two rows you can actually knock on a door for, and they're also the two rows that feed data into the ones you can't.

The metric that actually runs this whole system: save rate

If you remember one number from this post, make it save rate. Not streams. Not followers. Save rate.

A stream can happen passively, through autoplay, while you're half-listening to something else. A save means someone stopped and decided they wanted the song again. MusicPulse's breakdown of save-rate data found that tracks hitting a save rate above 15% in their first week were roughly 4.7 times more likely to get picked up by Spotify's recommendation systems than tracks sitting under 5%. That's not a rounding difference. That's the gap between a song that disappears after release week and one that keeps circulating months later.

What counts as healthy? A good save rate for a single runs roughly 3% to 5% of streams, and a song with 10,000 streams and a 5% save rate is actually in better shape than one with 50,000 streams and a 1% save rate. You can pull both numbers straight off Spotify for Artists: saves divided by streams, times 100. I've written before about which five numbers on that dashboard are actually worth your attention, and save rate is at the top for exactly this reason. If you've never checked it for your own release, stop here and go look. It'll tell you more than your monthly listener count ever will.

The actual steps, in order

None of this is a hack. It's just the sequence that respects how the algorithm gathers evidence about your track in the first place.

1. Nail your metadata and genre signals before release. When Spotify sees your track repeatedly co-occurring with specific similar artists, it gets more confident recommending you to those artists' fans through Discover Weekly and Radio. That starts with correct genre tags, a clean mix, and a release that actually sounds like the lane you're claiming. Say indie-pop in the metadata while the mix sounds like bedroom lo-fi, and you've confused the content-based layer before it even got a chance to help you.
2. Build a pre-save push so the first 48 hours aren't flat. Spotify watches engagement from day one. A pre-save campaign through your socials, your email list, or a landing page gives the track a concentrated burst of saves right at release, which is exactly the window the algorithm weighs heaviest.
3. Get placed on real playlists next to artists who sound like you. This is the step that actually feeds collaborative filtering. The same logic that applies to editorial placements applies to smaller, curated independent playlists too: listeners who are already primed for your sound are more likely to save, not just skim past. This is the piece of the chain we build campaigns around at PlaylistGrow. We've run 2,665 campaigns for 2,388 artists since 2019, mostly in Rap/Hip-Hop, Afrobeats, Pop, Dance/House, and R&B/Soul, resulting in 15,916 placements with 5,960 still active right now. It's the part of the funnel you can genuinely influence. It doesn't bypass the algorithm, it feeds it the right evidence.
4. Ask for saves directly, not just plays. When you post about a release, say the word. "Save this if you want it in your library" outperforms a vague "new song out now," because you're training your own audience to take the exact action that matters most.
5. Watch your skip rate in the first 30 seconds. Collaborative filtering partly runs on play counts and skip behavior, and content-based filtering checks whether the track's actual sound matches the cluster it's being tested against. If people bail in the intro, no save rate on earth fixes that. Film the hook three times if you have to. The first thirty seconds of the recording deserves the same fight you'd put into the first thirty seconds of a video.
6. Keep releasing. One great single followed by eighteen months of silence stalls the whole thing. New releases generate fresh data points that help the algorithm reassess your profile. A flat catalog with nothing new to evaluate just sits there.

If you're mapping this against an actual release calendar instead of reacting after the fact, our free 90-day release checklist walks through the pre-save, metadata, and playlist-pitch timing in the order it needs to happen, not the order most artists discover it by accident.

What doesn't move the needle, no matter how much people insist it does

Buying streams does nothing good here, and it can actively hurt you. The idea that more streams automatically means more algorithmic support ignores how the system weighs intent. If ten thousand people stream a track passively and almost none of them save it, the algorithm reads that as a track that doesn't stick. High skips with no saves can drag your positioning down instead of up. Playlist services that only promise raw stream counts, with no attention to whether those listeners are actually a genre match likely to save, are optimizing for the wrong number entirely.

Same goes for pitch spam. Submitting the same track to every curator you can find, regardless of genre fit, gets you streams from mismatched listeners whose low save behavior drags your ratio down rather than up. A smaller list of genuinely matched playlists beats a scattershot one every time, because the algorithm is grading fit, not volume.

A realistic timeline, not a promise

There's no fixed number of days or streams that flips a switch, and anyone citing one precisely is guessing. What the evidence does support: the window right after release carries outsized weight. Save behavior in the first one to two weeks feeds the collaborative filtering signal, a well-produced track with clear genre characteristics feeds the audio analysis, and an active, well-tagged artist profile feeds the NLP layer. Line those three things up around a release and Discover Weekly becomes a plausible outcome within a few release cycles, not a guaranteed one on any single song. If you want a structured push behind that release window instead of hoping the algorithm notices on its own, that's the exact gap our playlist promotion campaigns are built to close.

Nobody controls Discover Weekly directly. But almost every artist obsessing over it is ignoring the parts of the chain they actually can control: the metadata, the pre-save push, the playlist placements, the save rate they're asking fans for. Work those, and the algorithm does the rest on its own schedule, not yours. That's a less satisfying answer than a hack. It also happens to be the true one.