Shootout At the Fantasy Factory Part 2: What Should Streams Mean for Royalty Escalations?

As we noted in a prior post, Billboard reported on what might be called the “chart value” of streams compared to downloads.  To the extent that streams are counted in both the Billboard charts and the UK’s Official Charts, streams are valued at 100 streams to one permanent download.  This ratio was touted by Spotify at the recent artist meetings hosted by the Featured Artist Coalition in New York.

When you consider the royalty value of that ratio, however, those 100 streams are rewarded at a much, much lower penny rate than a download, even if you compare 100 streams to the $0.70 price of the lowest iTunes wholesale price of downloads.  As we noted previously, the 100:1 ratio implies a per-stream royalty of $0.007.

What About the Royalty Value?

Experience suggests that this seems high.  Quite high.  So just to confirm what Spotify representatives actually said at the New York meeting we revisited the meeting with Blake Morgan of ECR Music Group who was present and was quoted extensively in Billboard’s reporting.

Blake confirmed that Mark Williamson of Spotify told the room about the 100 to 1 chart value, but Mr. Williamson went even further.  According to Blake:

Mr. Williamson went on to explain that he felt the ratio should actually be more like 80:1 to properly and accurately reflect Spotify’s impact on the charts. He was being genuine, and said this while smiling…all in an effort to communicate how much Spotify matters, how it should matter even more, and how his company is beneficial to artists. He truly saw this as telling us good news. The only problem with this of course is that if he feels 80:1 is the accurate comparison of what streams are really worth versus permanent downloads, then how come artists don’t get paid the same for 80 streams as we do for one permanent download? That’s kind of an important detail to overlook, don’t you think?

I do think.  Solving for x with an 80:1 ratio and a typical $0.70 iTunes wholesale price (holding distribution fees at 0) the calculation would be a Spotify per stream rate of $0.70/80 or $0.00875, 25% higher.  I also asked Blake what his Spotify streaming royalties were on his last statement for comparison.  (You need to remember that Spotify, YouTube and others are not fixed penny rates but are calculated at a percentage of revenue basis–which is, of course, one of the problems with having no floor royalty rate.)

Blake was paid on two different rates for Spotify, one for the subscription service and another for the ad supported or “free” service.  Spotify has approximately 10 million paid users and another 30 million users of its ad supported service.  The 10 million paid subscriber number is a little vague itself as Spotify gives college students a 50% discount that it presumably passes along to copyright owners and there are probably other upsell/downsell variations on the rates.  Here’s what Spotify says about itself:

Free tiers:

Spotify has two tiers which are free to use: one for desktop and tablet computers and one for mobile.

The desktop and tablet free tier allows users to play any song in our catalog on-demand but users must view and listen to advertisements that interrupt their listening.

Users on our mobile (smartphone) free tier can only play music in “Shuffle mode”. They can not play songs on-demand or offline and face a number of other restrictions including limited song skips as well as being interrupted by advertising. For more information on our mobile free tier, please see this blog post.

Advertisers pay Spotify for exposure to users on our free tiers and in-turn fund the royalties that Spotify pays out for listening that occurs in these tiers.

Premium tier:

Spotify’s Premium tier gives users unlimited music across all of their devices including smartphones, tablets and TVs. Users can also temporarily download songs to their devices for listening on subways or airplanes, play music at the highest quality and they are never shown advertisements. This tier costs $9.99 per month.


Spotify’s total revenue comprises money received from advertising on the free tier and subscription payments on the premium tier.

So–the overwhelming majority of Spotify’s customers (about 75%) are using the free service–after four years of operating and “driving” free users to the premium services.  Wherever the free users are being driven, they’d taking the scenic route.

And if all streams are being counted at 100:1 or 80:1 ratio to downloads, it’s important to distinguish which pot the money is coming from for purposes of calculating the royalty value although Mr. Williamson did not.

Blake tells us that his per-stream Spotify royalty rate for the subscription tier is $0.005941864 and his royalty rate for the ad-supported tier is $0.000899194.  (The per stream rate should be more or less equal for all indie labels, although the total royalty payment will increase based on usage.  The per stream rate may vary depending on the deal and will almost certainly vary for major labels who will have higher per-subscriber minima and other goodies such as non recoupable payments that vary the effective royalty rate.)

Obviously, that’s a long way from $0.007 and even further from $0.00875.  And remember this–Spotify likes to say that they don’t pay artists directly because their deals are with labels.  In Blake’s case, Blake the artist is the same as Blake the label (as is the case for all artists on his ECR Music Group who all own their recordings).  So Spotify can’t run this dodge when it comes to Blake–Blake’s share is the label’s share.

When you consider that the artist royalty for signed artists is a fraction of the label’s share, now you understand why artists are freaking out over the minuscule streaming royalties they get from Spotify.  For example, if the label treated the gross as a royalty base price sale and paid at an album rate of 16% with no other reductions, you need to express the royalty rate in scientific notation.

Here’s what Spotify tells artists, which is of course not quite right, but close:

The “Spotify monthly revenue” box is not quite right because Spotify takes some off the top deductions before the allocation of your streams/all streams is calculated.  It looks something like this:

 “Approved Deductions” means: (a) with respect to the Ad-Supported Service only, any sales commissions or other fees paid to third person advertising agents directly relating to sales of Ad Inventory, such commissions and fees not to exceed in the aggregate thirty-five (35) percent of Spotify’s gross revenues from such sales in the US and Canada and twenty (20) percent in all other countries of the Territory; (b) with respect to the Ad-Supported Service and the Subscription Service, (i) any sales, use, value added or similar tax imposed on Spotify’s sales of Ad Inventory and the Subscription Service, and (ii) any fees paid to publishers or other interested parties pursuant to…the Standard Terms and Conditions (capped at 15%); and (c) solely with respect to the Subscription Service, any credit card processing fees or other transaction fees incurred by Spotify in connection with payments by Subscribers (capped at 3%).

The step where Spotify pays “70% to master & publishing owners” is of course incorrect, because there is no fixed percentage of Spotify’s gross revenues that is paid out.  (Pandora also runs this scam.)  Want the 70% to decrease?  Increase Spotify’s gross revenue from all sources.

Plus, the “artist payout” is expressed as the artist’s royalty rate x “70% to master and publishing owners”.  First you’d have to net out the publishing, of course, so Spotify’s own calculation is overstated by quite a bit-21% based on Spotify’s own explanation:

Spotify negotiates our royalty economics with labels and publishers in each territory where we operate. Our current payment agreements lead us to distribute (~)approximately 70% of our gross revenues to master recording and publishing rights (both mechanical reproduction and performance) holders. The precise division between these types of rights holders varies by territory in accordance with local laws and negotiated agreements. In the United States, for example, statutes dictate that publishers receive ~21% the amount that master recording owners receive….Recently, these variables have led to an average “per stream” payout to rights holders of between $0.006 and $0.0084. This combines activity across our tiers of service. The effective average “per stream” payout generated by our Premium subscribers is considerably higher.

Say what?

What is the Escalation Value?

It is very common for record deals, especially major label term recording artist agreements, to have an increase in the album royalty rate based on sales performance.  Aside from isolated instances, streams are never counted.  Why do you think that is?  Possibly due to the disparity in revenue?

Once upon a time, these major label royalty rate increases were accorded at 500,000 and 1 million units, but as those days are long, long gone, the increases (or “escalations”) are at much lower sales levels today.  Using our 16% album rate example, the escalation could be 1/2% or 1% at agreed upon sales levels, but this is not set in stone.

Since we’ve seen that Spotify believes their streams should be counted at an 80:1 ratio compared to downloads, even if they don’t want to pay royalties at the corresponding rate, it does seem that those streams should be counted for purposes of royalty escalations or the “escalation value.”  Because Spotify has so much involvement with major labels, we should expect that making such a proposal would not raise a single eyebrow.  Right?  Streaming is, after all, the inevitable future, right?  Right?

If Spotify is going to put their money where their mouth is, the chart value, royalty value and escalation value all need to true up.  And if they’re not going to true up, then let’s not con each other that these streams matter.

Spotify has no control over the chart value or the escalation value which involve third parties.  Spotify does have control over the royalty value.  However, Spotify doesn’t like per stream calculations:

Again, we personally view “per stream” metrics as a highly flawed indication of our value to artists for several reasons. For one, our growing user population might listen to more music in a given month than the month before (resulting in a lower effective “per stream”), while generating far more aggregate royalties for artists. As with any subscription service, our primary goal is to attract and retain as many paying subscribers as we possibly can, and to pass along greater and greater royalties to the creators of the music in our service. Theoretically, another service could generate higher effective “per stream” payouts simply by having users who listen to far less music.

But Mr. Williamson is more than happy to compare Spotify’s subscription tiers to download services when it suits him whether it’s 100 to 1, 80 to 1 or something else.

Here’s the problem that Spotify just does not want to grasp:  If you account to anyone for dozens to millions of uses of their property at a rate that requires a scientific calculator to compute they will always feel like they are getting screwed.

Because they are–if for no other reason because there is no floor on a per stream basis.  Getting that floor would at least be a start, especially if it maintained that 80:1 ratio.  And not only does Spotify have this problem, Apple is now floating trial balloons about cutting the Beats monthly retail subscription fee from $9.99 to $5 per month.  Enough is enough.

I’m sorry, but saying that you don’t like to talk about per stream rates because they are microscopic and humiliating really just doesn’t cut it.  Instead we get what freshman English students will recognize as glittering generalities:

We believe, however, that our service and the lives of artists will both be best if the World’s music fans enjoy more music than ever before in a legal, paid manner.

Well no kidding.  Like the man said, it’s a long way to the top if you want to rock and roll.

3 thoughts on “Shootout At the Fantasy Factory Part 2: What Should Streams Mean for Royalty Escalations?

  1. I submitted a comment on this a few days ago, but it seems to have vanished into the ether, so I will have another go.

    My main point was that there is something odd about Blake Morgan’s figures. Assuming that the average ratio of paid to free streams is 1:3, in line with the stated total numbers of users, Morgan’s figures would imply an average payout per stream, to an independent like himself, of only just over $0.002. This is much lower than the average payout to all rights-holders as stated by Spotify (between 0.006 and 0.008). It is also lower than other public data such as the much-quoted figures shared by Zoe Keating (about 0.0045.)

    Discrepancies like this need further investigation. One possibility, which I overlooked when I first attempted to comment, is that the ratio of paid to free may be much higher in actual streaming usage than in nominal user numbers. This makes a certain amount of sense. People who are actually paying for the service are likely to make sure they get their money’s worth, whereas people who just sign up to the free option will not necessarily use it much if at all. If this is the true explanation, then the payout to independents may not be quite as bad as may appear. But it does need investigating.


  2. 1. Please review our comment policy
    2. Review Spotify’s public calculation and note the mistake that overstates artist royalty base by including publishers.
    3. Note that the “average” from Spotify doesn’t distinguish which territories are included. Given that Spotify is subsidized in some countries it is important to distinguish which countries are included. Spotify doesn’t do this.


Comments are closed.