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

A curator I work with regularly told me flat out last month: "Artists keep asking me to get them on Discover Weekly. I keep telling them I don't have a button for that." She's right, and it's the single biggest misunderstanding in this whole business. Discover Weekly has no pitch form, no editor, no inbox you can slide into. Spotify assembles a fresh set of 30 songs for every listener, every Monday, entirely by algorithm. Nobody curates it by hand.

So the honest answer to "how do I get on it" isn't a hack. It's a chain of upstream events, and if you skip a link, nothing downstream happens. Your track needs to live on real playlists next to artists with a similar audience. It needs to earn saves and repeat listens from the people who hear it there. Only then does Spotify's recommendation system have enough signal to decide you belong in someone's Monday mixtape. This post is about building that chain properly, not finding a shortcut around it.

Why There's No One to Pitch

Editorial playlists like RapCaviar or New Music Friday are put together by human editors at Spotify, and you can submit to them through Spotify for Artists. Discover Weekly isn't that kind of thing. It's generated fresh for every single listener, which is a scale problem no team of editors could solve by hand even if they wanted to. There's simply no person on the other end to convince, no relationship to build, no email to send.

That's actually the more interesting story. Matthew Ogle, the Spotify product lead who built Discover Weekly, described the goal as wanting something that felt like a friend handing you a mixtape. The idea that made it workable wasn't some clever algorithm dreamed up in a lab. It was the sheer volume of playlists Spotify's own users had already built. Ogle told Music Ally ahead of the 2015 launch that user playlists, combined with other activity on the platform, capture people's deliberate curatorial intent and teach the system how songs relate to each other. That's the whole game in one sentence: your track needs to show up inside that web of human curation before the algorithm has anything to work with.

The Chain Nobody Explains Properly

Most advice about Discover Weekly jumps straight to tactics (release on a Friday! Get playlisted! Fix your metadata!) without explaining why any of it matters. So let's slow down.

Step one is collaborative filtering, and it's the backbone of the whole system. Spotify is constantly asking which listeners behave like each other. If a chunk of people who stream Artist A also stream Artist B, and a new listener only streams Artist A, Artist B becomes a candidate recommendation for that person. The more overlapping taste signals between two listeners, the stronger the confidence in that recommendation.

Step two is where your track has to physically exist inside that behavioral graph in the first place, and that's what real playlist placements do. Not any playlist. Definitely not a playlist stuffed with unrelated genres just to pad a follower count (curators call those junk drawers, and a lot of them get outright rejected before they're ever considered, for reasons we go into in why curators reject good songs). A track sitting on a well-curated, niche playlist next to three or four genuinely comparable artists teaches the collaborative filtering model that your listener overlaps with theirs. That's the entire mechanism. No placement, no overlap data, no candidate status.

Step three is what happens once people actually hear it there. Saves, skips, completion rate. This is the part artists have the most direct influence over, and it's where most campaigns get won or lost.

The Three Systems Working Underneath It

Spotify doesn't run one model. It runs three, layered together, and it helps to know what each one is actually listening for.

System	What it reads	What actually feeds it
Collaborative filtering	Overlap between listener behavior	Playlist placements, saves, follows, repeat plays
Audio analysis	The actual sound of the track	Mix quality, genre-typical structure, tempo and energy consistency
Text and NLP signals	What's written about you online	Blogs, articles, and metadata scraped from the web, weighted into relevance terms per artist, a process Harvard's Digital Innovation Lab has broken down in more detail

None of these can be gamed directly, which is worth sitting with because a lot of promo companies imply otherwise. A strong save rate in the first two weeks feeds collaborative filtering. A well-mixed track with clear genre characteristics feeds audio analysis. An active, coherent public profile feeds the text layer. You influence each of these by doing the work upstream, not by finding a backdoor into any one of them.

The Practical Steps That Actually Feed the Chain

1. Get onto real playlists before you worry about Discover Weekly at all

This is the step people skip because it feels like a detour. It's not, it's the entire on-ramp. Your track needs exposure to listeners who already stream music like yours, and a smaller, tightly-curated playlist next to genuinely comparable artists generates that overlap data far faster than a huge generic one. We've made the math case for this before in why a 2,000-follower playlist can beat a 100,000-follower one, and it's the same logic that applies here. Placement work is the core of what we do at PlaylistGrow, and how that's structured is here, aimed at getting tracks in front of the right listeners early, not at inflating a raw stream count.

2. Protect the first two to four weeks like they're the whole campaign

Spotify treats a new release as being in a kind of evaluation window right after it drops, and early behavior gets weighted more heavily than the same behavior months later. Front-load your promotion, your playlist pitching, and your fan outreach into that window instead of spreading it thin over months. If you want a structured way to plan for that window before you even release, our free 90-day release checklist walks through the pre-release steps that set it up properly.

3. Get your metadata and genre tags right, and be specific

This sounds boring. It isn't optional. A song tagged as indie folk will find different listeners than one tagged as experimental electronic, and vague or aspirational tagging just confuses the audio analysis layer about which taste cluster you belong in. Be accurate, not aspirational. Tagging your bedroom pop track as "alternative" because it sounds cooler doesn't help you. It puts you in front of the wrong listeners, who skip fast and hurt your signal.

4. Don't buy streams. Genuinely, don't.

Paying for streams doesn't help Discover Weekly placement and can actively hurt it. Artificial streams get excluded from the signals that feed recommendation algorithms in the first place, and detected artificial streaming carries real platform penalties on top of that. I've had artists ask me if buying a few thousand streams to "get the ball rolling" is a reasonable shortcut. It isn't. It's noise the algorithm is specifically built to filter out, and if it's flagged, you're digging out of a hole instead of building momentum.

5. Collaborate deliberately, not randomly

If you collaborate with an artist whose listeners overlap with your target audience, you get access to their listener network for exactly the purposes Discover Weekly cares about. A feature verse or a co-write with an artist one or two tiers above you, in a genuinely adjacent lane, does more for your collaborative filtering profile than almost any other single move, because it creates instant, legitimate overlap in listening behavior.

About Those Save Rate Numbers Everyone Quotes

Ask five different sources what save rate you "need" for Discover Weekly and you'll get five different numbers. Boost Collective's own breakdown of save rate versus skip rate puts country, indie, and rock tracks at roughly 5 to 7% saves on average with lower skip rates, while hip-hop and electronic tracks tend to run lower, largely because those genres see more passive, playlist-background listening. Other write-ups put the "trigger" threshold as high as 20%. Nobody outside Spotify's own data science team knows the precise number, and honestly it almost certainly moves by genre, release size, and even by week.

Treat any specific percentage you read, including the ones above, as a rough compass, not a target to hit exactly. What's consistent across every source is the direction: saves and completions matter far more than raw stream count, and a skip in the first thirty seconds reads as a negative signal no matter whose benchmark you trust.

Where This Fits Into the Bigger Picture

We've run 2,163 campaigns for 1,951 artists since 2019, most of them in rap and hip-hop, Afrobeats, pop, dance, and R&B, and made 12,778 playlist placements, with 5,551 still active today. None of that is a Discover Weekly guarantee, and I'd distrust anyone who tells you otherwise. What that volume of placements has taught us is the pattern above, though: tracks that land on the right niche playlists and hold listeners past the first thirty seconds are the ones whose artists come back a few months later saying their algorithmic reach picked up on its own. That's not us pulling a lever. That's the chain working the way it was built to.

You don't get on Discover Weekly. You build the conditions where Spotify's own systems decide you belong there, real placements next to the right artists, genuine saves in the first few weeks, honest metadata, and patience with a system that was never built to be pitched in the first place.