Skip to main content

Streaming Services Propose New Royalty Rates for Songwriters and Publishers

Spotify, Amazon and Pandora are proposing lower rates for publishers and songwriters for the period spanning 2023–2027 than any year since 2018.

Spotify, Amazon and Pandora are proposing lower rates for publishers and songwriters for the period spanning 2023–2027 than any year since 2018.

On Thursday (Oct 21), the CRB posted the digital music services’ rate proposals for the term, also known as Phonorecords IV, and those streamers are recommending lower rates, mathematically speaking, than any of the yearly figures set by the CRB Phonorecords III determination that covers 2018–2022. That rate is, however, currently under an Appeals Court remand and it’s unclear if its previous determination will stand.

Spotify, Pandora and Amazon are harkening back to the Phonorecords I and II rate determinations that covered 2008–2012 and 2013–2017, respectively, and generally using those rates as guidelines in their new proposals. As for Apple Music, it is essentially proposing whatever rates the CRB judges decide in the remand process for Phonorecords III, and basically the same rate formula from that determination but with fewer wrinkles.


The National Music Publishers Association (NMPA) is already on record condemning the digital services for its rate proposals, which it claims will lower the percentage of streaming revenue paid to songwriters and publishers if enacted.

The Digital Media Association (DiMA) issued a statement about the rate proceedings on Thursday emphasizing that growth in streaming’s  popularity will benefit everyone.

“I believe the answer lies not with any rush to judgment or allusions to war, but in truly grappling with this cognitive dissonance — ever-growing revenues for rightsholders, billions of dollars invested into catalogs, and new tools and features that help bring more music to more fans than ever before in a highly competitive landscape, alongside genuine frustrations by creators,” DiMA CEO Garrett Levin said in a statement. “How do we make modern music economics work for everyone? That should be our focus — preserving long-term industry growth and ensuring that it benefits as many people as possible.”

For Spotify’s premium paid subscription — the company’s main revenue generator — the service is proposing a roll-back to the headline 10.5% service rate enacted in the Copyright Royalty Board Phonorecords I and Phonorecords II rate determination. That would be below the headline rates enacted during Phonorecords III, which began at 11.4% in 2018 and escalates annually to 15.1% in 2022 and is is currently under an Appeals Court remand.


That 10.5% of service revenue would be measured against a second bucket, the larger of which would emerge as the all-in pool from which the performance royalties paid to the PROs were subtracted, leaving the mechanical royalty costs for the services.

The second bucket would be created in a two-step process: The first step would involve multiplying 80 cents per subscriber and measuring it against 21% of total content cost, i.e., what’s paid to the record labels. Whichever of those two revenue pools was lower would then become the second bucket.

That 21% of content cost, as well as the 80 cents per subscriber, are the same rates from Phonorecords I and II — but the content cost is lower than the under-remand Phonorecords III rate that began at 22% of content cost, escalating annually to 26.2% of content cost in 2022.

In all, Spotify offers rate proposals for nine service tiers created by the CRB process down through the years — even though Spotify doesn’t actually offer all of them. For most tiers, Spotify proposes a 10.5% of revenue and 21% of content cost as components of the rate formula. For ad-supported streaming — the next largest tier after premium paid subscriptions — Spotify is proposing a 10.5% of revenue against 22% of what’s paid to labels — the same rates used in Phonorecords I and II — and whichever is greater becomes the all-in-bucket from which the performance royalties are subtracted. Again, those rates are lower than what the CRB judges determined for the under-remand Phonorecords III of 12.5% of revenue and the escalating content cost, which in 2021 is 25.2%.

Pandora’s rate proposals are very similar to Spotify’s.


Amazon proposes a 10.54% of revenue all-in headline rate across the board, versus 80 cents per subscriber for its mobile paid subscriber tier. The percentage rate is considerably lower than this year’s current rate of 14.2% of revenue for a paid tier, which is scheduled to go up to 15.1% of revenue next year, although that overall rate and rate structure is up in the air due to the remand. The introduction of a per-subscriber fee at this point in the rate formula is a new wrinkle.

For its non-portable devices, which generally includes subscribers accessing by computer but in Amazon’s case also includes lower-priced plans tied to its smart speaker devices, Amazon is proposing the greater of an all-in rate of 10.54% of revenue versus 19.1% of the total content cost (i.e., what it pays to record labels).

For the above three tiers (portable, non-portable and ad-supported), the resulting bucket would be an all-in revenue pool from which the performance revenue paid to performance rights organizations would be subtracted and whatever is left over is the mechanical rate pool. That pool would then be divided by the total number of streams for the period to determine the per-stream rate, which is then paid out to each song in accordance with the number of streams that song generated during the month.

For its pared-down service available to all Amazon Prime subscribers, which offers a limited inventory of songs, the company is proposing a straight $0.00085 per-play rate.

But just as important as the rates for the three tiers that are based on 10.54% of service revenue, Amazon is looking to reduce that revenue pool by also asking for a deduction of the fees it pays to the third-party app stores.


The Apple App Store, for example, charges a 30% fee on monthly subscription payments for subscribers who signed up for the Amazon service using an Amazon app downloaded from the App Store. The Google store also charges the same type of fee. Subscriptions that come through the app stores, however, generally carry a premium price of $12.99, so this reduction would move the subscription price down to about $10, or the normal rate.

Amazon is also asking for another deduction of up to 10% of revenue for undefined payments to third parties, outside of what it wants to deduct for payments to app stores. This too is a new deduction, and it is unclear what third-party payments it could apply to.

Adding up the two new deductions — 30% and 10% — doesn’t, however, mean that the service revenue pool would be reduced by 40%. Those deductions would only apply to those subscriptions that come in through the app stores or in some way involve other third-party charges.

Spotify is also proposing new deductions from service revenue — some nuanced and some more obvious, like for podcasts and other non-music content.

Overall, these new wrinkles in the rate formula, and what gets counted as revenue, would likely reduce the overall size of the all-in 10.54% of revenue bucket in two ways: a lower rate, applied against a smaller pool.

Apple, meanwhile, wants to adapt the Phonorecords III rate formula but simplify it by eliminating the step that was previously calculated by applying a percentage against whatever is paid to the record labels.


Instead, Apple would have an all-in bucket — including the total publishing payment of both performance and mechanical royalties — calculated by applying the headline percentage of revenue rate against service revenue, which, as previously mentioned, is currently at 14.2% of revenue but subject to remand. After the performance payments to the PROs are deducted, whatever is left would be measured against a 50 cents per-subscriber floor. Whichever is greater would then be the mechanical pool.

In addition, it sounds like Apple is planning to offer a new subscription model, as it’s proposing a new limited-offering service category for full-catalog services with substantially limited functionality as compared to premium services. That would have a lower mechanical floor of 25 cents per subscriber, rather than the more commonly applied 50 cents per subscriber. The lower per-subscriber rate would be to reflect the service’s reduced functionality, although it’s unclear how it would accomplish that since it’s offering a full catalog.

Apple is also proposing an introductory hardware bundle that would have a fixed monthly mechanical minimum of 33 cents per subscriber for up to two years, after which the standalone royalty formula would kick in.