How to Actually Get on Discover Weekly (There's No Pitch Form, So Now What)

Discover Weekly has no pitch form, no submission box, no editor listening to your track and deciding it's good enough. That trips people up, because almost everything else in the Spotify ecosystem, at least the parts artists try to influence, runs through a human somewhere. Editorial playlists have curators. SubmitHub and Groover have actual people approving or rejecting submissions. Discover Weekly has none of that. It's a machine running collaborative filtering, and once that sinks in, most of the mystery around it clears up fast.

The mechanism, in plain terms

Spotify groups listeners into taste clusters based on what they play, save, and skip. Then it serves each cluster songs it thinks they haven't heard but are statistically likely to enjoy. Your song enters that machine when real listeners in a cluster near your music start saving and finishing it, usually because it showed up next to comparable artists somewhere first: a playlist, a Radio session, a Release Radar drop. No saves, no signal. No signal, no Discover Weekly.

That's the whole chain. Every tactic that actually moves the needle feeds one link of it. Everything else is noise.

Why "how do I get on it" is the wrong question

The question assumes Discover Weekly works like editorial playlists, where someone says yes or no. Editorial playlists like RapCaviar or New Music Friday are curated by people, and you pitch those before release through Spotify for Artists, a process we've broken down step by step elsewhere. Discover Weekly runs on a different system with a different logic. It isn't judging your song. It's testing what listeners do with it after the fact.

Editorial placement can still help, indirectly. A track that lands on an editorial playlist and earns saves from those listeners strengthens the collaborative filtering signal, which makes Discover Weekly pickup more likely in the following cycles, according to Music24's breakdown of the process. So pitching editorial isn't wasted effort. It's just aimed at a different target that happens to feed the one you actually want.

The actual chain, step by step

This is the part most "how to" posts skip, so let's slow down.

1. Your neighbors matter more than your numbers

Collaborative filtering works by proximity. If listeners who already stream an artist adjacent to you start finishing and saving your track, the system infers the rest of that cluster might too, and starts testing your song on them, per AndR's rundown of the trigger metrics. That's why context beats volume. A thousand streams from listeners nowhere near your genre teach the algorithm nothing. A hundred streams from people already playing three artists that sound like you teach it a lot.

This is also why playlist placement matters, though not for the reason most people assume. It's not the raw play count a playlist sends you. It's repeated co-occurrence, your song showing up around the same artists over and over, until the system draws a line between them. Get placed on curated or independent playlists next to two or three genuinely comparable artists, and you're feeding this exact link. It's the piece we work on at PlaylistGrow, and it's why our campaigns are built around genre-matched curators instead of the biggest playlist that'll say yes regardless of fit.

2. Saves and completions do the talking

Once new listeners hit your track, what they do next is everything. Boost Collective's research on save rate versus skip rate found save rate is the stronger predictor of algorithmic playlist success, while skip rate acts more like a quality filter that can suppress reach. A skip in the first thirty seconds reads as a bad match. A save reads as deliberate intent, and that carries far more weight than a passive stream ever will.

Most artists get this backwards, in my experience. They chase stream count because it's the number that feels good to watch climb. But Chartlex's review of the platform points to save rate and completion rate as the two indicators that still carry the most weight for Discover Weekly and Release Radar. A smaller, more engaged audience beats a bigger, indifferent one every time.

3. Spotify tests before it commits

Your song doesn't get dumped into every possible Discover Weekly on day one. Chartlex describes it as a trial run: a track might appear in a few hundred Discover Weekly playlists first, and if those listeners save, complete, and return to it at above-average rates, distribution expands to thousands more the following week. If they don't, it quietly stops trying. Nobody flags you, nothing crashes. The song just fades out of the test pool, which is why so many artists swear nothing happened at all.

What actually feeds each link

• Get your metadata right. Genre and mood tags tell Spotify which cluster to test your song against. Spaceloud's guide is blunt about this: be specific, since a song tagged indie folk finds different listeners than one tagged experimental electronic. And don't over-engineer the title or artist name trying to game search. Spaceloud also notes Spotify penalizes over-optimization, so it can work against you.
• Activate your real fans first, fast. The first 48 to 72 hours set the tone for how the system reads the rest of the song's life. Get your actual audience saving the track before you spend a dollar on strangers. Email list, close-friend Instagram broadcast, a pinned link, whatever gets people who already like your sound to click save on release day.
• Use Release Radar as the on-ramp. It's the most accessible algorithmic playlist for emerging artists, since it just needs someone to already follow you, per Tunebump's notes on popularity scoring. Engagement there is often the first real data Spotify has before it'll test you anywhere colder.
• Land on curated playlists next to comparable artists. This is the association-building step from above, and it's a slow burn, not a one-time event. Keep pursuing third-party curator placements in your specific lane rather than the biggest playlist that'll take you.
• Space your releases out. A single every four to six weeks gives the system a steady stream of fresh data to test. Drop once a year and you're asking the algorithm to start over from zero every time.

If you're planning a release and want to line up the pre-save, the fan activation window, and playlist outreach ahead of time instead of scrambling release week, our free 90-day release checklist walks through that timeline.

The numbers everyone quotes (and why I'd ignore most of them)

Every music blog has a magic number. Here's a sample of what's floating around right now, side by side, so you can see how little they agree.

Source	Claimed threshold
Independent playlist promo blogs	10,000+ streams, 4,000+ listeners, 400+ saves within 28 days
Algorithm-focused promo blogs	20,000+ total streams plus a 3%+ save rate from cold Discover Weekly listeners specifically
Popularity-score guides	An internal Spotify popularity score of 30 or higher
Save-rate focused guides	A 20%+ overall save rate, independent of raw stream count

None of these are official. Chartlex says it plainly: these aren't published Spotify thresholds, Spotify doesn't release exact numbers, but they correlate with algorithmic pickup in the data these companies track. That's an honest way to frame it, and I respect a company that admits the limits of its own data. Boost Collective's framing is the more useful one: treat these as directional guidelines, not hard gates. Chase the behavior, saves, completions, repeat listens from the right people, not a specific number you saw on somebody's blog. The threshold moves by genre, by how niche your sound is, by how Spotify happens to be weighting things that month.

How long this actually takes

Slower than people want. MusicPulse's coverage of Luminate's 2025 mid-year streaming report found the median time from release to first Discover Weekly placement, for independent tracks that eventually reached 50,000+ streams, was around 11 days. That's for songs already headed somewhere. For most independent artists, one great single isn't the trigger. Orphiq's research puts meaningful Discover Weekly placement more commonly after three to five consistent releases with strong engagement from the right audience, which tracks with what we see running campaigns. It's rarely one song that flips the switch. It's a run of releases that keep building the same signal.

What actively hurts you

Two things, and both are avoidable.

First, fake engagement. Bots and streaming farms might bump a number for a week, but Spotify has gotten aggressive about catching it, and Boost Collective notes that getting flagged can suppress your tracks from algorithmic playlists entirely. Not a slap on the wrist. A suppression across the board.

Second, promoting to the wrong audience at scale. Blasting your track to a broad, untargeted list feels productive. It isn't. Boost Collective's guide points out that promoting to a broad, undefined audience often produces high skip rates, which tells Spotify the song isn't resonating. The algorithm doesn't punish you on purpose. It just reallocates attention elsewhere, which in practice looks exactly like being punished.

Where we actually fit into this

We're not going to pretend we can get you on Discover Weekly, because nobody can sell that honestly. Nobody controls it, including Spotify's own editorial staff. What we do is work the part of the chain that's actually reachable: getting tracks placed on curated, independent playlists next to genuinely similar artists, so the association-building step has real data to work with. Since 2019 we've run 1,336 campaigns for 1,238 artists and made 7,562 playlist placements, 4,672 of which are still active, mostly across rap and hip-hop, Afrobeats, pop, R&B, and electronic. That's the part of the funnel we know cold.

What Spotify's algorithm does with that after the fact is genuinely out of anyone's hands, ours included, and I'd be lying if I told you otherwise.