A friend of mine put out a song two years ago. It did well — not famous well, but genuinely well. Playlists picked it up, a couple of small sync placements happened, and the counter on her dashboard eventually rolled past one million streams. A million. That's a number you'd expect to change someone's life.
It bought a used car. A decent one, but a used one.
When she told me this, she wasn't bitter about it. She was confused. She had assumed, like most people do, that somewhere in a server room there was a line of code that read payout = streams × rate, and that if she could just find out what the rate was, she could do the math and plan her year. That line of code does not exist. The number she got wasn't a rate times a count. It was her share of a pot, and the pot was divided by rules almost nobody outside the industry can explain.
Here's what those rules actually are.
There is no per-stream rate, and that's the whole story
Every calculator you find online will tell you streaming pays somewhere around $0.003 to $0.005 per stream. Those numbers aren't lies exactly — they're averages people worked backward from real payout statements. But treating them as a rate is like calculating your "per-mile cost of living" from your annual salary and your odometer. The math works. It just doesn't describe any actual mechanism.
Spotify itself is explicit about this: it does not pay royalties according to a per-play or per-stream rate. What actually happens is a pro-rata pool. Each month, in each market, the platform takes its subscription and ad revenue, sets aside its own cut, and puts roughly the remainder into a pot for rights holders. Then it counts every stream that happened in that market that month. If your songs were 0.001% of the streams, you get 0.001% of the pot.
Your payout isn't a price per stream. It's a slice of a pie whose size you don't control and whose other eaters you've never met.
This has a consequence people find deeply unintuitive: your earnings can fall while your streams rise. If your plays go up 20% but total platform streams go up 30%, your share of the pool shrank. You did better and got paid less. Nothing went wrong; the model just doesn't work the way the word "rate" implies.
It also means the geography of your audience matters enormously. A stream from a paying subscriber in a high-revenue market contributes far more to that market's pool than a free-tier stream in a market with low ad rates — commonly quoted as a four-to-eight-fold difference. Two artists with identical stream counts can receive wildly different checks purely because of where their listeners live and whether those listeners pay.
The 1,000-stream cliff
In April 2024, Spotify added a rule that reshaped the bottom of the market: a track must reach at least 1,000 streams in a rolling 12-month window before it earns anything at all. There's also an undisclosed minimum number of unique listeners, kept secret specifically so nobody can game it by looping their own song on twelve devices.
Fall below the line and your track earns zero. Not a small amount — zero. And the money that would have gone to those tracks doesn't vanish; it's redistributed to the tracks that cleared the threshold.
The platform's argument is that these micro-payments were mostly never reaching artists anyway. A track earning $2.40 a year often couldn't clear a distributor's minimum withdrawal, and the administrative cost of moving those pennies exceeded the pennies. Better, the reasoning goes, to consolidate that money where it can actually be paid out.
The counter-argument is loud and specific. In a December 2025 survey of labels, 65% reported significant revenue losses from the change and another 20% reported a slight negative impact, with 92% saying they strongly opposed the threshold. Critics point out the pain isn't evenly spread: it lands hardest on classical, jazz, regional, and non-English catalogs — genres with deep back catalogs of individually low-streaming tracks. A label with 4,000 recordings each doing 400 streams a year used to earn something from all of them. Now it earns nothing from any of them.
Both things can be true. The pennies genuinely were hard to distribute, and the fix genuinely took money from the long tail and handed it to the head. That's not a paradox — it's just what redistribution looks like when you're on the wrong side of the line.
Where the money goes before it reaches a person
Here's the part that explains my friend's used car. The pool payment doesn't go to the artist. It goes to rights holders, and an artist is usually not one, or is only partly one.
A single stream generates two separate royalty streams that flow through different pipes:
| Right | Who owns it | Who typically collects |
|---|---|---|
| Recording (master) | Label, or the artist if independent | Label → distributor → artist |
| Composition (publishing) | Songwriter and publisher | Publisher / collection society → songwriter |
The recording money — the larger share — goes first to whoever owns the master. On a traditional label deal, that's the label, which recoups its advance and recording costs before the artist sees anything. An artist can be "in the black" on streams and still owe money against an unrecouped advance. On an independent release, the artist keeps far more of the master share but paid for everything up front and takes a distributor cut off the top.
The composition money splits again between songwriter and publisher, and if a track has five co-writers, that half-share becomes a fifth of a half-share. This is why you'll see a song with enormous stream counts producing a modest check for the person who sang it — they may have written none of it and own none of the recording.
So when someone says "a million streams pays about $4,000," the honest version is: a million streams sends roughly $4,000 into a system, and how much of that reaches any individual human depends on contracts they signed years ago.
The numbers at the top look nothing like the numbers in the middle
Spotify paid out more than $11 billion to the music industry in 2025 — the largest annual payment to music from any single retailer in history, growing more than 10% year over year while other industry income sources grew closer to 4%. Independent artists and labels accounted for about half of all royalties. Those are real, large, genuinely improving numbers.
And in that same year, 1,500 artists generated over $1 million, while more than 13,800 generated at least $100,000.
Sit with the second set for a second. Thirteen thousand eight hundred artists earning six figures is a lot of working musicians — far more than the old CD economy supported at that level. It's also a vanishingly small fraction of the millions of accounts uploading music. Both the "streaming is paying out record sums" story and the "almost nobody can live on it" story are supported by the same spreadsheet. They're just answering different questions: how much money is in the system, versus how many people it reaches.
The pie got much bigger. The number of forks grew faster.
What this means if you actually make music
I'm not going to tell you to quit. But the model has clear implications, and they're mostly about where you aim rather than how hard you work.
Depth beats breadth. Because payouts scale with a listener's revenue contribution and tracks must clear a threshold, 1,000 people who play your album repeatedly on paid accounts are worth more than 50,000 people who each hear one song once in a playlist and never return. Chasing playlist placements that generate passive, one-time plays optimizes for the metric that pays least.
Own what you can. The single largest variable in what reaches your bank account isn't your stream count — it's what percentage of the master and publishing you control. A deal that doubles your streams while halving your ownership is not obviously a good deal. Read the recoupment clause specifically; that's where the used car goes.
Treat streaming as the top of a funnel, not the funnel. For most working artists, streaming income functions as discovery and validation, while the actual money comes from live shows, merchandise, sync licensing, teaching, session work, and direct-to-fan platforms. The artists doing well aren't usually the ones who cracked the algorithm. They're the ones who converted streams into a relationship they can monetize somewhere else.
Register everything. Publishing royalties go uncollected constantly because nobody filed the paperwork with the right society. That's free money left in an account somewhere with your name on it, and it is genuinely common.
The short version
Streaming royalties aren't a price. They're a share of a monthly revenue pool, split by your fraction of total plays in each market, weighted by whether your listeners pay and where they live. Tracks under 1,000 annual streams earn nothing, and that money is handed to tracks above the line. Whatever comes out is then divided between recording rights and publishing rights, and only after contracts and recoupment does anything reach a person.
Understanding this doesn't put more money in anyone's account. But it does replace a frustrating mystery with a solvable problem: instead of asking "why is the rate so low," you can ask the questions that actually move the number — who owns my masters, where do my listeners live, and what am I building that streaming can point toward.
My friend still has the car. She also has a mailing list of 3,000 people now, and last year they were worth considerably more to her than the million streams were. That's not a sad ending. It's just a different map than the one she started with.
Figures cited are as of writing and change over time; royalty terms vary by platform, market, and contract. This is general information, not financial or legal advice.


