The advice comes back around like a chorus. A service raises its monthly price, the timeline fills with cancellation threats, and someone from the industry side says the calming thing: a higher subscription price means a bigger royalty pool, a bigger pool means bigger checks, so anyone who cares about artists should be rooting for the increase. It is the closest thing streaming economics has to folk wisdom, and it carries the considerable advantage of being mostly right.
Mostly. The distance between "mostly right" and "right" is where every serious argument about payouts actually lives, and it is not a rhetorical distance. It is arithmetic you can run in a spreadsheet in an afternoon. What follows is that arithmetic: where the rule holds, the four places it leaks, a procedure for checking any announced increase yourself, and a more honest version of the rule to reach for the next time someone corners you about it at a conference bar.
Do higher subscription prices actually raise artist payouts?
Yes, and by less than the headline suggests. Recorded-music licences are built mostly as a share of the service's revenue, so when revenue rises the money owed to rightsholders rises with it, automatically, without anyone renegotiating anything. What does not rise automatically is what an individual track earns, because the same pool is divided by a stream count that has been growing faster than prices have. A price rise makes the pie bigger. It does nothing about the number of forks.
That is the whole piece in five sentences. The useful part is knowing which of those sentences breaks first under load.
Where the rule holds up
The reason the folk wisdom works at all is structural, and it is worth being precise about it because the precision is what makes the rest of the argument tractable.
Major recorded-music deals are typically written as a "greater of" calculation. The service owes the largest of several prongs: a negotiated percentage of the relevant revenue, a per-subscriber minimum, and in some deals a floor tied to plays. Publishing sits on top of that with its own share, set by negotiation in some territories and by statutory rate proceedings in others, and it is generally also expressed against revenue rather than against a fixed per-play figure.
Two consequences follow, and both are load-bearing.
First, the licence cost is not an exogenous input that arrives from outside and ruins the quarter. For the percentage prong, it is a function of the service's own top line. Revenue up, rights payments up, in near lockstep, by construction. A service that raises prices is choosing to raise the amount it owes.
Second, the per-subscriber minimum is the prong that bites when revenue per user is low. Discounted plans, student tiers, and multi-user products can push a service toward the floor rather than the percentage, which is why heavily discounted growth is expensive growth and why services are structurally motivated to raise prices even when they would rather not spend the churn. When the minimum binds, the pool does not shrink proportionally with a promotional price. It sits on the floor.
So: bigger revenue, bigger pool. That part of the advice is not spin. It is the contract.
The four places the rule leaks
Nobody pays the sticker price
The number in the press release is the standard individual plan in one market, usually the largest one. It is not what the average subscriber pays.
Average revenue per user is dragged down by family and duo plans, student rates, free trials, prepaid and gift codes, telco and device bundles sold at wholesale, and promotional windows that run for months. Increases tend to land on the individual tier first and on other tiers later, or not at all. Bundle partners are on their own contract terms and do not reprice because a press release said so.
The practical effect is that a headline increase of a given percentage moves blended revenue per user by a meaningfully smaller percentage, and it moves it with a lag as the base rolls over. If you model a price rise by applying it to the whole subscriber base on day one, your number will be wrong in the direction that makes you look naive in front of a CFO.
The denominator does not hold still
Pro-rata allocation, still the default nearly everywhere, divides the pool by total qualifying streams and pays out share of consumption. It has one obvious property that is easy to state and hard to sit with: every additional stream in the system dilutes every other stream.
Consumption has grown steadily as more subscribers listen for more hours across more surfaces. Supply has grown faster. And this is the point where a publication about AI music tools has to be honest about its own subject matter. The marginal cost of putting another finished-sounding track into the world has fallen off a cliff. Generative tools, functional-audio farms, sleep and focus catalogues, and white-label ambient libraries all deposit into the same denominator as a record that took nine months and a string section. Pro-rata was designed in a world where making a release cost something. That world is gone, and the allocation model has not caught up to it.
So the pool can grow at a healthy clip while blended per-stream drifts flat or down, and both facts can appear in the same quarterly report without anyone lying.
The pool is not the payee
Between the money leaving the service and the money landing on an artist statement sit a distributor or label share, a producer, a publishing split routed through a different pipeline entirely, foreign exchange, a possible advance still recouping, and increasingly a monetisation threshold: at least one major service now requires a track to clear a minimum annual stream count before it accrues at all, with the withheld money redistributed into the same pool.
All of those terms were fixed at contract signature, often years before the price change. An artist on a deal signed in a different decade experiences a price rise through a filter built for different conditions. When they report that the increase did not show up on their statement, they are usually describing their contract accurately, not misreading it.
Not all subscription revenue counts as music revenue
The percentage prong applies to a defined revenue base, and definitions are negotiable. When a service adds non-music content to a subscription and reclassifies the product as a bundle, the share of that subscription attributed to music can be recalculated under the bundling provisions in the relevant agreements and rate structures. Consumers see one price. The revenue base underneath it can be a different shape than it was last year.
This is the leak that catches analysts out most often, because it does not show up as a price change at all. Nothing on the checkout page moved.
Price-check any increase in six steps
This takes about an hour with public filings and a spreadsheet. Do it once and you will never take a payout press cycle at face value again.
-
Start from revenue, not from price. Pull reported subscription revenue for a period and average subscribers across the same period, then divide. You should see revenue per user that sits well below the sticker price of the standard plan, often by a wide margin. If your figure is close to the sticker price, you have used period-end subscribers instead of the average, or mixed a global revenue line with a single-market user count.
-
Chart that figure across eight quarters against the price-change timeline. You should see a line that is visibly flatter than the price history. Mix eats increases. The gap between the two lines is the leak, quantified, and it is the single most useful chart in this whole exercise.
-
Turn revenue into a pool. Apply the combined share going out across recording and publishing. Use a range rather than a point estimate; the widely circulated figure of roughly two-thirds of revenue leaving for rights is an order of magnitude, not a spec sheet, and it varies by market and product. You should see modelled rights cost that tracks revenue almost exactly. If it does not, look for a reclassification or a shift in market mix.
-
Find the denominator. Use reported total streams for the same service and period if it is published, or back into a blended rate from your own catalogue's payouts divided by its streams on that service. You should land in fractions of a cent, and you should see it differ materially by market. A single global per-stream figure is a fiction that averages a high-ARPU market and a low-ARPU one into a number describing neither.
-
Re-run the model with the increase applied only where it actually landed. Weight by the subscriber share of the affected plans and markets, and phase it over the renewal cycle rather than switching it on at once. You should see the pool move by a fraction of the headline percentage. That fraction, not the headline, is the number to quote.
-
Compare pool growth to stream growth across the same window. If streams grew faster than the pool, blended per-stream fell even though every absolute number in the earnings release went up. That is the tell, and it is the ordinary result rather than the scandalous one. It is also the entire reconciliation between "record industry revenue" and "my statement went down."
The claim, and where it gives out
| The claim | Roughly right because | Where it breaks |
|---|---|---|
| Higher prices mean a bigger pool | Rights payments are a negotiated share of revenue, so they scale with the top line | Revenue per user rises by much less than the sticker price, and with a lag |
| A bigger pool means a higher per-stream figure | More money is being divided | The denominator grows too, and supply growth is not constrained by production cost |
| Per-stream is a rate the service sets | Some deals contain per-play floors and per-subscriber minimums | For most payouts it is an output of a division, not an input anyone chose |
| Licensing costs are outside the service's control | Rates are negotiated with counterparties who drive hard bargains | The largest single driver of the bill is the service's own revenue |
| If artists earned less, the service kept the difference | Something clearly did not reach the artist | Most of the gap is allocation, mix, contract terms, thresholds and recoupment |
"Rising licensing costs" is a true sentence doing dishonest work
Which brings us to the messaging, and to the part that is genuinely irritating rather than merely complicated.
When a service attributes a price increase to what it pays for content, the statement is accurate. It is also the least useful true thing available, because it invites the customer to picture a greedy counterparty extracting more money for the same product. What it obscures is that the bill went up largely because the service decided its revenue should go up. The percentage prong did what percentage prongs do.
The defensive framing is a small choice with a compounding cost. Trained over enough cycles, it teaches consumers that music is an expensive input rather than the product, and it teaches artists that the service views them as a cost centre. Neither party ends up better disposed toward the business.
There is a version that is both honest and shorter. We are raising the price because the product is worth more than it costs. A majority of the increase flows straight to the people who made the music, because that is how our agreements are written. Here is the ratio.
No one has to be blamed in that sentence, and it happens to be the strongest argument the category has. A monthly subscription that covers effectively the entire recorded canon remains cheap against almost any comparable media bundle, and it has stayed cheap through a long stretch of general price inflation. That is a defensible position, held plainly. Reaching for the cost excuse forfeits it.
What actually moves the per-stream number
If the goal is a higher figure per play rather than a bigger absolute pool, price is not the main lever. Five things matter more.
Market mix. Growth concentrated in high-ARPU markets raises blended revenue per user without any price change at all. Growth concentrated in emerging markets does the opposite while looking excellent in the subscriber column.
Superpremium and tiering. Higher-priced tiers move revenue per user faster than an across-the-board increase, because they are opt-in and the increment is larger. Whether the associated content commands a different royalty treatment is a deal question, and the answer varies.
Allocation reform. Artist-centric variants, monetisation thresholds, functional-audio reclassification and fraud enforcement all change how the pool is cut. Be clear-eyed here: these redistribute rather than grow. They raise per-stream for the tracks that remain qualifying by removing claims from the denominator. That may be the right policy, but describing redistribution as growth is how the industry loses the next round of trust.
Denominator hygiene. Removing artificial streams and reclassifying noise uploads does more for a mid-tier artist's rate than a modest price rise does, and it does it faster.
Renegotiated minimums. When the per-subscriber floor moves in a renewal cycle, the effect on discounted and bundled tiers can outweigh anything happening on the retail page.
The honest version of the rule
Keep rooting for price rises. They are the cleanest available mechanism for growing the pool, they are long overdue against inflation, and the contracts guarantee that the majority of the increment leaves the building. Stop promising anyone that it will be visible on their statement.
The rule, rewritten to survive contact with a spreadsheet: a subscription price rise reliably grows the pool and unreliably grows the cheque. The pool is a pricing question. The cheque is an allocation question, a mix question and a contract question, and those are the three fights that are actually worth having.
This week, run the smallest version of the check. Take one service, one market, four consecutive quarterly statements for a catalogue you control. Chart your blended per-stream against that service's reported revenue per user for the same four quarters, on the same axis. If your line went down while theirs went up, you now know precisely which of the four leaks you are standing in, and you can stop arguing about the headline price entirely.
The price of a subscription sets the size of the pool; everything anyone actually cares about happens after the division.
Not sure which tool to use?
Compare the top AI music and sound tools side by side — honest reviews, real pricing, no sponsorships.
