The Chart That Spotify Can't Play

TL;DR: Half the albums on the TOP 20 Albums Worldwide are not on Spotify. The #1 album earned roughly $3,350 from fewer than 200 purchases — the Spotify equivalent of 840,000 streams. The paid-playlist industry charges $285 for a chance at earning $1.50 back. The artists who skipped all of it are at the top of the chart.

---

The search that started this

I had not opened Spotify in about a year. The TOP 20 Albums Worldwide had just closed its second weekly edition, and I wanted to check something simple: how many of our 20 charting albums were on the platform that every distribution guide, every music marketing course, every “how to release a single” blog post treats as the centre of the universe.

I started at #1. Mouth Ulcers, Silent Pictures. Not there.

Number 2: Massive Attack, Boots on the Ground. Not there either — but something else was. A track by someone called NEONPH4NTOM, with AI-generated cover art, with the words “Boots on the Ground” and “Massive Attack” printed on it. A lookalike, squatting where the real thing should be. I stared at it for a while. I had been away for a year, and the place had got worse.

Number 3, Steve Wynn. Not there. Number 4, Weatherglow Collab Project. Not there. Number 5, 2XT. Not there. I kept going. By the time I reached #20, the count was 10 out of 20. Half the best-selling independent albums in the world, in a given week, do not exist on Spotify.

The chart did not go looking for this. It counts sales — money changing hands, filtered by an anti-gaming algorithm. The chart closed at midnight on Friday. The 10 absent albums include Massive Attack, one of the defining electronic acts of the last 3 decades. Steve Wynn, who co-founded The Dream Syndicate in 1981 and whose The Days of Wine and Roses is the album critics point to when tracing a line from the Velvet Underground through the Pixies to Nirvana. Black Milk, who signed on as Slum Village’s in-house producer as a teenager and has since worked with J Dilla, Pete Rock, and Jack White. Lunar Shadow, whose The Pall Of A Past World charted #1 in both Metal Hammer Germany and Deaf Forever.

Real careers. Real audiences. Real sales. No Spotify.

190 wallets and 840,000 ghosts

The numbers are modest. That is the point.

Mouth Ulcers at #1 sold fewer than 200 copies — digital and physical combined — for roughly $3,350. A debut EP from a London post-punk quartet who describe their music as “for vampires to dance to.” At Spotify’s per-stream payout of $0.003 (the Duetti 2024 Music Economics Report pegs it at $3 per 1,000 streams, the lowest among major services), matching that $3,350 would require approximately 840,000 streams. For a band whose biggest gig was probably a London basement last month. They do not have 840,000 streams. They do not have a Spotify page. They have fewer than 200 people who heard something and paid for it, and that was enough for #1 worldwide.

At #20, Inferi’s Heaven Wept sold just under 100 copies for roughly $1,400. Inferi are actually on Spotify. Their most popular track has accumulated about 3.5 million lifetime streams — over years. One week of direct sales earned the equivalent of 10% of those years of streaming. And album sales do not stop after the first week, even if you leave the chart. At $6 a month — a single sale every few weeks — an artist earns more than the literal nothing Spotify pays to 62 million of its 100 million tracks that fall below the 1,000-stream annual minimum.

The ratio: 1 purchase equals roughly 4,000 to 5,000 streams. One wallet replaces 4,000 pairs of earbuds.

Average transaction across the chart: $14 to $18. Higher than digital album pricing. A meaningful fraction of those buyers are ordering vinyl, ordering CDs, ordering the thing they can hold. Each of them is worth 3,600 to 4,400 plays. The streaming model needs thousands of people to produce what 1 person with a credit card produces.

Ninety-seven copies is a Tuesday night at a small venue. A mailing list built over 6 months of gigging. Any working musician can picture selling 97 of something. Very few can picture 350,000 streams.

What experienced superstars do when the goal is money

Massive Attack are not strangers to Spotify. Mezzanine is there, Blue Lines is there, Heligoland is there. They have an artist page. They know exactly how the platform works.

For Boots on the Ground — a charity single with Tom Waits’s first original song in 15 years, profits split between the ACLU and the Immigrant Defense Project — they chose Bandcamp. Released on Thursday, the final day of the chart window. 1 day of trading. By midnight Friday, 174 people had bought it. Roughly $3,020.

On Spotify, $3,020 in a single day would require approximately 755,000 streams between midnight and midnight. Drake does that. Massive Attack, 3 decades of global recognition, probably would not. But 174 people buying a 12-inch on Bandcamp? Done before the day was over.

This was not a distribution delay. This was a choice — made by an act with the infrastructure and the experience to release anywhere they want. When the goal was money, not visibility, not playlist adds, not algorithmic favour, they went where money is. And it was not Spotify. If a sign were needed that Spotify is not a source of income, Massive Attack holding a charity single off the platform and raising $3,020 in a day from 174 people is the sign, written in neon, at the top of the chart.

Maybe they will release it there eventually. Or maybe they are at the point every experienced act reaches — the point where you do not need the most expensive discovery machine in the industry, because the people who care already know where to find you.

The lookalike at #2

Back to that NEONPH4NTOM search result. Massive Attack’s actual single is not on Spotify. What is on Spotify is a track with AI-generated cover art, the band’s name and single title written on it, from an account in Pakistan, released in the same window. A fan searching for the real thing finds the fake first and the real thing never.

Spotify introduced Artist Profile Protection in March 2026 to combat exactly this. It is happening to Massive Attack anyway.

The contamination is not small. Deezer’s AI Radar — the only publicly transparent detector among major streaming services — tracked AI uploads rising from roughly 10,000 per day in January 2025 (10% of daily uploads) to 60,000 per day by January 2026 (39%). Beatdapp, the fraud-detection firm partnered with major labels, estimates at least 10% of all global streams are fraudulent, driven primarily by organised criminal operations laundering money through streaming platforms. Their co-founder told Billboard that when they started, they assumed fraud ran 1 to 3%. They now put it at 10%, with industry-wide losses at $2 to $3 billion annually.

An Ipsos survey for Deezer found 97% of listeners could not distinguish AI-generated tracks from human-made ones in a blind test. The AI-generated “psych-rock band” The Velvet Sundown reached 1.4 million monthly listeners in mid-2025, appearing in users’ Discover Weekly feeds and sitting on mood playlists alongside The Beatles and Billie Eilish before admitting it was not human. AI tracks have landed on the verified Spotify pages of Toto and the deceased Texas songwriter Blaze Foley. Here We Go Magic’s Luke Temple called it “so predatory, and so terrible.”

The platform the industry recommends as the default destination is simultaneously missing half the best indie music, hosting AI-generated fakes in its place, and running on a stream economy where $1 in $10 may be fraudulent. The artists who are not there are not missing anything.

The $285 toll booth on the road to $0.003

If streaming paid badly but cost nothing to enter, the case against it would be weaker. But an entire intermediary industry has grown between artists and their fractions of a cent, and it charges real money for access to imaginary returns.

SubmitHub: $1 per credit, roughly 6,000 curators, 10 to 20% approval rate in practice (3 to 8% for niche genres). Groover: €2 per pitch, 10 to 12% placement rate, 2 to 8 weeks before your track is rotated out. Playlist Push: $285 minimum, $450 average, no refunds once curators begin reviewing, no guarantee of any placement. Stream-guarantee services like Indie Music Academy and Boost Collective: $79 to $1,950 for placement into closed networks, a model that sits uncomfortably close to what Spotify explicitly prohibits.

None of these sell a placement. They sell a chance. A curator who accepts 100 submissions per month on Playlist Push at the $5 midpoint clears $500 in fees regardless of whether they add a single song. Playlist owners earn nothing from Spotify for third-party tracks on their playlists. The fees are the business. Not the curation. The fees.

Feed a realistic campaign through the arithmetic. 5 placements averaging 500 followers, 2,500 follower-impressions, 1 to 5% play rate from followers who did not choose the song. Best case: 2,000 streams, $6 in royalties on a $285 spend. Realistic case: 300 to 800 streams, $1.50. Worst case: below the 1,000-stream threshold, $0.

One Trustpilot reviewer: “I spent $280 and this gave me 150 streams in total, so I paid $2 for a stream. You could literally send someone to ask people on the street if they would listen to this song for $2 and achieve the same or more.”

Ari Herstand, whose Ari’s Take is the most widely read independent musician advice blog, retracted his recommendation of Playlist Push after 3 test campaigns. His funk-soul song was placed on playlists including “Boybands and Hit Songs,” “2000 Indie Rock,” and “Ares 2008-2012 Electronic Dance Music.” His description of the experience: “This was so traumatizing that I asked George to remove all of our songs from all the playlists we got.”

Andrew Southworth spent several thousand dollars on playlist promotion in 2019 and watched his monthly listeners crash from 15,000 to under 1,000 after each campaign ended. He switched to Meta ads and now publishes head-to-head data showing paid playlists deliver under 1% save rates while Facebook conversion campaigns hit 26%. The paid playlist audience — people who did not choose the song — generate the worst possible engagement signals. Spotify weights save rate, skip rate, and completion. Paid streams that do not convert do not merely fail to help. They teach the algorithm to show the song to fewer people. One Boost Collective customer put it precisely:

“You will get a bump in streaming numbers at the expense of wrecking your algorithm, making it impossible for anyone who would actually like your music to find it, ever.”

The same reviewer reported being banned from Spotify after the campaign.

The curators, the bots, and the 91 Swedes

Side-Line, the Belgian music site, ran a 2023 investigation in which curators on submission platforms openly offered to sell placement for $49 to $280 without listening to the submitted track. Multiple Spotify Community threads from 2023 and 2024 document playlist operators adding tracks from roughly 2,000 artists per day without permission, bot-streaming them, then upselling “promo” when the victim artist notices the spike. Innocent artists caught in Spotify’s subsequent fraud dragnet have lost their accounts for streams they neither bought nor understood they were buying.

In September 2024, the Southern District of New York indicted Michael Smith of Cornelius, North Carolina. He ran 1,040 bot accounts streaming roughly 636 AI-generated songs per day, peaking at 661,440 daily streams. Total fraud: over $10 million. He pleaded guilty in March 2026 and forfeited $8 million. His words from the indictment: “We need to get a TON of songs fast to make this work around the anti fraud policies these guys are all using now.” His partner at an AI music company, in 2019: “what we’re doing musically here… this is not ‘music,’ it’s ‘instant music’ ;)”

The collateral damage on real artists is extensive. Jonah Baker, a 150-million-stream artist earning roughly $200,000 a year, received distributor strike notices without having paid any promotion service. The rapper Viper had a track removed for “100% artificial” streams even though it sat on official Spotify editorial playlists. Herstand’s own album Brave Enough was pulled in 2017 after a service he paid turned out to be running click-farms. His conclusion: “Because Spotify can’t figure out how — or just don’t care — to go after the actual perpetrators, they go after the victims. The artists.”

And underneath all of this sits Spotify’s own contribution. Liz Pelly’s January 2025 Harper’s investigation, drawn from her book Mood Machine, documented that since 2017 Spotify has operated a programme called Perfect Fit Content. An internal team called Strategic Programming commissions stock-music-style tracks from partner companies — Epidemic Sound, Firefly Entertainment, Hush Hush LLC, among others — and seeds them across 150+ chill, ambient, jazz, sleep, and focus playlists. Brian Eno was removed from “Ambient Chill” and replaced by commissioned PFC. Jon Hopkins was removed. Bibio was removed. A fabricated artist called Ekfat, complete with a fake Icelandic biography about the nonexistent “Reykjavik music conservatory,” racked up 4 million streams on a single track. Dagens Nyheter uncovered 91 Swedes operating behind 13,000 pseudonymous Spotify tracks. A former Spotify employee told Pelly: “Some of us really didn’t feel good about what was happening. It’s just not fair. But it was like trying to stop a train that was already leaving.”

The economic logic is plain. Commissioned buyout tracks carry no per-stream royalty obligations. Every PFC play is pure margin. The platform that tells independent artists to pitch for editorial placement has been replacing real artists with house-brand pseudonyms on its own playlists, in the exact genres — ambient, jazz, classical, focus — where patient, demanding music most needs support.

The discovery myth and the 6-minute tone

The standard defence of all of this is that Spotify is not about revenue. It is about discovery. Being findable. Being in the machine that recommends you to strangers.

MIDiA Research surveyed 10,000 consumers in September 2025. The study was called “All Eyes, No Ears,” which gives away the finding. Of the listeners who had heard music on social media in the previous month, 48% had not subsequently streamed any of it. Of the 20% who followed an artist on TikTok after discovering them, only 26% actually listened to more of that artist’s music. A third of social-discoverers admitted they had simply forgotten the song’s name by the time they might have looked it up. The youngest cohort — 16 to 24, the generation the industry claims lives on playlists — converts discovery to continued listening at the lowest rate of any age group. Just 19%. Virality and fandom are different things, and they are getting more different.

Edison Research’s 2024 Music Discovery Report found word of mouth reaches 82% of U.S. listeners, up 14 points since 2016. Growing in the algorithmic era. Nielsen’s trust research has repeatedly found 88 to 92% of global consumers trust recommendations from people they know above all other forms of communication. The most powerful discovery mechanism in music predates Spotify by the entire history of human speech.

Then there is live. A 2024 study in the MEIEA Journal: 68% of concertgoers discovered at least one new artist through an opening act, 73% actively sought information about openers after buying a ticket. Eventbrite and MusicWatch’s data: live-discovered fans spend 20 times more than non-ticket-buying consumers. Twenty times. Luminate: a median 42% lift in local streaming during concert week, across roughly 1,000 shows. Live events drive streaming more than streaming drives live.

In a previous Signal & Noise piece, I described discovering Brìghde Chaimbeul by a poster outside a 500-year-old venue in Stirling, Scotland. A physical piece of cardboard. 1 sentence on my phone about Scottish smallpipes. A gig in a 166-seat auditorium. A CD bought at the merch table, signed by the artist. None of this involved a playlist. All of it involved attention — the sustained, focused, undistracted kind that a streaming playlist structurally cannot produce.

One of Chaimbeul’s tracks is 6 minutes of a single tone. I am not sure it contains distinct notes. On a Spotify playlist, it would be skipped in seconds — and the data says so. Paul Lamere, former director of Spotify’s Echo Nest, published the numbers: 24% of Spotify plays are skipped within the first 5 seconds. 35% within 30 seconds. Nearly half abandoned before the track ends. At a gig, you are focused. You learn to enjoy it. You start to understand it. Music — especially indie music, the kind that fills this chart, Dreamcastle’s LEGO-catalogue vaporwave and DJ Reckless’s sub-bass that ships with a printed disclaimer about speaker damage and Castle Rat’s medieval metal performed in handmade chainmail — needs understanding. It is not bubble gum. The format has to permit the music to work. The skip button does not permit anything.

Ohio State musicologist Hubert Léveillé Gauvin studied Billboard Top-10 hits from 1986 to 2015 and documented a 78% collapse in average intro length, from roughly 20 seconds to 5 seconds. Chartmetric’s 2024 data shows the average Spotify-charting song now runs about 3 minutes, down 30 seconds since 2019. Producers told Billboard plainly: “I have producers in the studio this week just going through and making songs shorter.” The platform is not a neutral surface. It is a selection pressure, and it selects against the music that rewards patience. The Mouth Ulcers EP that sits at #1 on the chart does not need to survive that pressure. It sold on Bandcamp, where nobody skips anything, because you paid for it.

The heretical advice

Everything that follows goes against every distribution guide, every “how to release a single in 2026” blog post, and every piece of advice from companies whose revenue depends on artists uploading to every platform simultaneously, 4 to 6 weeks before release, with pre-save campaigns and Spotify Canvas videos and playlist pitches queued up like prayers.

Consider the incentive structure behind that advice. Then ignore it.

Do not put your album on streaming services. Not immediately. Not in the first months after release. The album is yours. The frequencies are yours. The tracklist, the sequencing, the mastering choices, the decision to open with a 6-minute drone or a 30-second punk blast or 4 bars of silence — those are yours. The moment you hand everything to a streaming platform, you lose control of how it is consumed, you surrender scarcity, and you hand the most engaged period of your release cycle to a system that will pay you $0.003 per listen while the fans who would have bought it stream it for free. You traded a $17 purchase for four-tenths of a cent.

And the grip tightens. Once your album is on the platform, the platform commands: conform to our loudness standards, keep intros short, make sure the hook lands before the 30-second royalty threshold, structure your tracklist for completion rate, release singles on a schedule that feeds the algorithm. Your deliverables become their deliverables. Your art bends toward their metrics. Your 6-minute tone becomes a liability.

Instead: release 1 song on streaming services. A teaser. Let it share the album art with the actual album. Think of it as the streaming edition — shaped for the attention economics and technical requirements of those platforms, so that the album itself does not have to conform. You are the master of your frequencies. Let people hear what you sound like. Let them know something larger exists. Do not give the rest away.

Then create a landing page. 1 URL. It leads everywhere: the streaming single across all services, the full album for purchase on Bandcamp or your own store, the physical product, the merch. The digital album preview shows more tracks than the streaming single — proof that buying the whole thing is worth it. Since roughly half of people buying vinyl and a significant percentage of CD buyers do not own a player, some bonus tracks can sit there with a wink — playable only on actual hardware — giving people a reason to own both the digital album and the physical object.

Vinyl has grown for 19 consecutive years. RIAA data: U.S. vinyl revenue hit $1.4 billion in 2024, outselling CDs for the 3rd straight year. The Vinyl Alliance surveyed Gen Z buyers and found 76% purchase at least once a month, 80% own a turntable and actively listen. These are not nostalgic collectors. These are fans making effort-heavy purchase decisions — the polar opposite of background-stream behaviour. The committed-fan economy has relocated outside streaming’s gravity, and it did it while nobody in the advice industry was looking.

The artists who skipped all of it

King Gizzard & The Lizard Wizard have released 27 studio albums and 73 live albums, most Bandcamp-exclusive. Their “bootlegger” programme releases masters and artwork free to fans and indie labels for physical pressing. An entire parallel vinyl-label economy grew out of it.

Sault operates in radical anonymity. No interviews. No photos. No videos. They dropped 5 albums simultaneously as free downloads. Made 1 album available for 99 days, then erased it from the internet. Sold out their first live show on £99 tickets with zero marketing.

Sleep Token built a Platinum-certified following and sold out 2026 arena tours through cryptic lore drops and private listening sessions — the antithesis of algorithm-friendly consistency.

Billy Strings made Highway Prayers the first bluegrass album in 22 years to debut #1 on Billboard’s Top Album Sales chart. The audience was built on Grateful Dead–style touring devotion, not playlist placement.

Godspeed You! Black Emperor have released 2 official band photographs in 30 years. Given a half-dozen interviews. Credit their 16mm film projections as a band member. Refuse formal encores. 5 post-reunion albums on Constellation Records, all selling to consistent critical and commercial success.

None of them needed a playlist. All of them needed an audience willing to pay attention. As it turns out, those are different things entirely.

The TOP 20 Albums Worldwide — where Mouth Ulcers sit at #1 with a debut EP, where Massive Attack raised $3,020 for the ACLU in a single day, where 2 Dreamcastle LEGO vaporwave records hold adjacent chart positions and Inferi’s Nashville death metal closes out the list — is the only place where all 20 albums can be heard on a single page. Not because we planned it that way. Because the other platform is missing half the music.

Fewer than 200 people bought the #1 album. Just under 100 bought #20. These are not large numbers. They are real ones. Streaming metrics are large numbers. They are also, increasingly, imaginary — padded by bots, diluted by AI, gamed by curators whose business model is the rejection fee, valued at a fraction of a cent each, and policed by a fraud system that punishes the victims when it catches the crime. The paid-playlist economy is a reverse subsidy: artists pay $200 to $1,000 per campaign to push their work through infrastructure populated by hobbyist fee-collectors, AI-generated filler, and platform-commissioned ghost artists named after fictional Icelandic conservatories.

190 real transactions. 840,000 imaginary ones. The industry spent 15 years pretending that was a difficult choice.

Fuck them.