The Stream Pays Two-Thirds to the Music. Most of That Stops at Universal.

You stream a song on the train to work, and somewhere a meter quietly turns. Most people picture the payout the way the press describes it: a coin worth about four-tenths of a cent, dropping into the artist's pocket. That coin doesn't exist. Spotify itself says so - it pays no fixed per-stream rate; what you generate is a sliver of a monthly pool, sized by your track's share of every stream on the platform.⁴ But the more important fiction is what happens to the money after it leaves Spotify. It does not flow to the artist. It flows to a holding company in the middle of the music, and most of it stops there. That company, more often than not, is Universal Music.

The official story is that streaming finally pays the musicians - that two-thirds of every dollar goes to the people who make the music. The number is real. The conclusion is wrong. Two-thirds goes to rightsholders, which is a very different sentence, because a rightsholder is usually a label, and the label is the toll gate the money has to pass through before any of it becomes an artist's.

Follow the dollar all the way down

Start with a dollar a Premium subscriber's listening generates. Spotify keeps roughly a quarter to a third of its revenue and distributes the rest to rightsholders.⁶ Of what flows out, about 80% goes to the recording side and 20% to publishing.⁶ On the recording side, the master owner - almost always the label - receives around 80% of the royalty value.⁵ Now the crucial layer, the one the '70% to artists' framing hides: the label does not hand that to the artist. Under a typical major-label deal, the artist's contractual cut is 13–20% of the label's share - which works out to roughly 10–16% of the total stream value, while about 64% of that recording-side share lodges with the label.⁵ That is the whole machine in one sentence. The artist is not the recipient of the stream. The artist is a line item the label pays out of money it already controls.

And there is a trapdoor beneath even that 10–16%. The artist's royalty is not cash; it is a credit against the advance the label paid up front. Until that advance is recouped, every stream simply pays down a debt the artist owes - the royalties accrue, but no money moves.⁵ So the real flow is narrower still: a slice of a slice, often withheld entirely until the books balance. The label collects from the first stream. The artist collects, if ever, from the millionth.

Why the label sits where it does - and won't move

None of these percentages are written into law. The recording-side royalty rate is freely negotiated between label and platform - and the three majors, Sony, Universal and Warner, sit across that table with far more leverage than anyone else.⁶ Universal alone represented over 3.2 million recorded tracks at the close of 2023.³ A streaming service cannot offer a credible catalog without the majors, and a single major can withhold a slice of the world's most-streamed music. That asymmetry is the moat. The label is not the most creative party in the chain and not the most loved one - it is the unavoidable one. It owns the masters, it negotiates the rate, and it sets the artist's cut. Each of those is a tollbooth, and Universal owns all three at once.

The scale this produces is hard to overstate. UMG turned over €11.1 billion in 2023, more than a billion dollars a month, on roughly €2.37 billion of adjusted EBITDA.¹² Streaming generated 54% of first-half 2023 revenue, and subscription streaming - the high-quality stuff - grew 12.8% at constant currency.¹³ This is what a toll on other people's creativity looks like at industrial scale: the company that makes none of the songs sits at the gate through which nearly all of its catalog must pass.

The reform that flows the money even more toward Universal

In September 2023, Universal and Deezer announced what they called the first comprehensive 'artist-centric' streaming model, pitched as a fix for a system that pays musicians too little. Read the mechanics, not the press release. 'Professional artists' - defined as those with 1,000+ monthly streams from 500+ unique listeners - get a 'double boost' in royalty weight; a play someone actively searched for gets boosted again; and white-noise and non-artist filler, about 2% of Deezer's streams, is demonetized entirely.⁷ The pool of money does not grow. The weighting changes - and it changes in favor of catalog that clears Universal's own bar. The reform's stated effect was to lift payouts to professional artists by roughly 10%.⁷ That lift comes from somewhere: from the emerging and casual artists who fall below the threshold, and from the demonetized noise. It is not redistribution to artists broadly. It is a redrawing of the pool toward the artists Universal already earns the most from.

And notice the venue. The pilot launched in one country, France, with one label - Universal - on a service of about 10 million subscribers, against Spotify's hundreds of millions.⁷ As a redistribution program for the world's musicians, that is a rounding error. As a proof of concept for a weighting system Universal helped design and would like every platform to adopt, it is a deliberate first move. By the time UMG reported its full-year 2024 results, management was citing 2024 as the first full year of implementing artist-centric royalty models across its DSP partnerships.⁹ The pilot was never about France. It was about the template.

The fair objection: isn't the label earning its cut?

The honest counter is that the label is not a parasite - it takes real risk and does real work. It pays the advance before a single stream exists, fronting the cost of recording, marketing and distribution against an artist who may never recoup. Most signings lose money; the few hits carry the rest. By that logic, the 64% is the price of capital and a portfolio bet, not pure rent. There is truth in it. But the truth has limits. The risk argument justifies a premium; it does not obviously justify owning the masters in perpetuity, setting the artist's rate from a position of overwhelming leverage, and then proposing the 'reform' that tilts the next dollar further toward the catalog you already monetize best. A risk premium is defensible. A self-dealt re-weighting of everyone else's money is the part the manifesto exists to obscure.

There is one more wrinkle the label can't fully control: the structural ceiling on easy growth. UMG's ad-supported streaming revenue grew just 3.6% at constant currency in 2023, while subscription revenue, the higher-quality tier, rebounded to 8.2% growth in 2024 - but the ad-supported pool, where most casual listeners live, remained essentially flat.¹⁹ When one part of the pool stalls, the only way to keep the total flow rising is to change how it's divided - which is exactly what 'artist-centric' does. The reform isn't generosity arriving in good times. It's the choke point optimizing its share before the easy gains run out.

Universal Music makes its money the way a lock on a canal does: it doesn't grow the music, it doesn't move the boats, it simply owns the one gate every stream of its catalog must pass through, and it sets the level of the water. Two-thirds of the dollar reaches the rightsholders - and the rightsholder, more often than not, is the gate. The genius was never a clever royalty rate. It was standing in the place every play has to pass, owning the masters, setting the artist's cut, and now writing the rules of redistribution itself - so that even the reform meant to help the artists flows, quietly, the other way.