Investigation · Economics

There is no per-stream rate

Name: There is no per-stream rate
Published: 2026-08-12
License: https://creativecommons.org/licenses/by/4.0/

Published answers to the most-asked question in music disagree with each other by more than twenty times. The interesting part is why that is not mostly dishonesty.

We collected published per-stream figures for the major platforms and measured the spread between them. For YouTube, two published answers to the same question differ by a factor of 21. Rather than pick a side, this piece explains the mechanism that makes every one of those figures unquotable — and states the question that can be answered.

21×Spread between the lowest and highest published YouTube figure

~86%Share of Spotify tracks reported as affected by the 1,000-stream eligibility threshold

$0The true per-stream rate for any track below that threshold, whatever the published figure says

The rate is an output, not a price

What the published figures actually say

Platform	Lowest published	Highest published	Spread
YouTube / YouTube Music	$0.0007	$0.015	21×
Tidal	$0.008	$0.015	1.88×
Apple Music	$0.006	$0.010	1.67×
Spotify	$0.003	$0.005	1.67×
Amazon Music	$0.004	$0.006	1.50×

A 1.5× spread might be rounding. A 21× spread is a different question being answered under the same headline — in that case, partly YouTube's ad-supported video platform and YouTube Music's subscription service being reported as one thing.

Why no figure can be correct in advance

A vinyl turntable and retro radio player — **A play used to be a physical act with a fixed price. Now it is a share of a pool that has not been divided yet.** Photo: Firzafp · CC BY-SA 4.0 · source

The major services do not pay a price per play. They pool the period's subscription and advertising revenue and divide it by each rights holder's share of total streams. The per-stream number is therefore an output of a division that has not happened yet. It cannot be quoted ahead of time because it does not exist until the period closes, and it moves when anyone else's listening moves.

That single mechanism explains most of the spread:

Who listened matters. A premium subscriber's stream is worth roughly two and a half to three times an ad-supported one, so a blended average conceals the thing that decides your number.
Tidal and Apple have no free tier. Every stream comes from a paying subscriber. Their higher published rates are largely a composition effect, not a more generous deal — a distinction almost never made in the comparison tables.
Gross or net is rarely stated. Before or after the distributor's cut is a different number, and both get published as "the rate".
Territory is rarely stated. Subscription prices differ enormously by market, and so therefore does the pool per stream.

And the figure that overrides all of them. Since 1 April 2024, a Spotify track must reach 1,000 streams in the preceding twelve months to be eligible for royalties at all. Reporting indicates roughly 86% of tracks fall below it. For those tracks every published rate is wrong by the same amount, because the real rate is zero. Spotify's position is that 99.5% of streams occur on tracks above the threshold — which is true and is a different claim from 99.5% of artists.

The question that can be answered

"What does a stream pay" has no stable answer. "What determines my share of the pool" does, and it is rarely asked because it does not fit in a headline:

What proportion of your listeners are paying subscribers rather than ad-supported.
Which territories they are in, because the pool per stream follows the local subscription price.
Whether your tracks clear the eligibility threshold at all.
What your distributor takes before the money reaches you, and whether it is a percentage or a flat fee.
Whether you also hold the publishing side, which is paid separately and through entirely different bodies.

Every one of those is knowable in advance. None of them appears in a per-stream table.

What this is not

This is not an accusation. Most of the sources publishing these figures are doing something reasonable: readers ask for a number, so a number is supplied. The structural problem is that the honest answer — "the question is malformed" — is not something you can put in a marketing headline, and most publishers in this space have something to sell. We do not, which is the only reason this page is easy for us to write.

What would prove us wrong

A platform publishing a fixed, guaranteed per-stream price. That would make the rate an input rather than an output and this analysis would be obsolete.
An audit of real statements showing the published ranges are accurate and stable across tiers and territories. We hold no statement data and would defer to anyone who does.
The spread narrowing to rounding error across a larger sample of sources, which would mean the disagreement is a small-sample artefact rather than a structural one.

Limits, plainly. These are published claims, not measurements. We audited no royalty statement and hold no payout data of our own. Sources may be measuring different things — gross or net, one tier or blended, one territory or global — and that is precisely the finding rather than a flaw in it. Collected on a single day; published figures are revised often and quietly. A wide spread does not prove any source is wrong.

The data

Every published figure, the computed spreads, the structural facts and the stated limits ship as an open dataset.

Download data.json How the pool works

Collected 2026-08-12 · CC-BY 4.0, commercial reuse permitted, attribution and a link back required.
Referenced: Spotify Loud & Clear · Spotify track monetization eligibility · Spotify for Artists royalties guide · The MLC
Related: Per-stream rate · Pro-rata vs user-centric · Distribution fee · Who answers the money questions