Category Archives: Streaming

Parsing Spotify’s Financials

Spotify recently filed its annual report with regulators in Luxembourg, where the company is registered. The full annual report isn’t available online, but there’s enough data in various articles to put together a reasonable picture of its financials. As the poster child of streaming music (and its largest beneficiary), Spotify’s financials are illustrative of the state of streaming music overall, and so it’s worth looking closely at them to discern trends.

If this topic is of interest, you should also listen to a recent episode of the Beyond Devices Podcast, in which I interviewed Ryan Wright, the CMO of Kobalt, a music startup which is streamlining the process of getting payments from services like Spotify to labels and artists. We discussed Spotify specifically during the episode, and Ryan had an interesting take on the free/paid split. You may also find this column I wrote for Variety on the RIAA’s recent annual report interesting.

Massive growth

We already knew to expect massive revenue growth, given Spotify’s periodic updates on subscriber numbers, but the financials certainly bear this trend out powerfully. The chart below shows annual revenues, and you can clearly see the much bigger jump in 2015, relative to any previous year:Spotify revenue growthThe interesting thing here, of course, is that 2015 is also the year Apple Music launched. Far from putting a dampener on Spotify’s growth, Apple Music seems to have accelerated it, if anything. More likely, the streaming music industry has simply reached a tipping point, and both the Apple Music launch and Spotify’s success in 2015 are symptoms of that common cause.

Worsening losses

With all that growth, and the increasing scale that comes with it, one might expect that Spotify would be inching closer to profitability. However, that wasn’t the case, as it actually went backwards in 2015, in that both operating and net losses grew rather than shrinking (though margins were up slightly):Spotify marginsThe company still loses money, and one of the biggest reasons is that the vast majority of its revenue goes straight to paying the labels for the music it streams. And, instead of that number falling as a percentage of revenues, it’s risen for the last two years:Spotify royalty costs as percent of revenueThat’s problematic, because it means that if Spotify is ever to make money, it has to squeeze its other costs harder to generate a profit. Interestingly, though Spotify has often said that labels get a roughly 70% cut of its revenues from streaming, this contribution is much higher – in the mid 80s. That appears to be because Spotify has a whole set of complex and interrelated agreements with labels to pay minimum amounts regardless of usage, and as such its actual payouts are higher than that standard 70% cut in at least some cases.

The good news is that Spotify’s other cost lines are shrinking as a percentage of revenue, even as they grow in real terms. Personnel costs were just 13% of revenues in 2015, down from 18% three years earlier, but external consulting fees were still 2.5% of revenues in 2015, while advertising and PR made up another 4.5%. Given that Spotify only has about 16 points of gross margin to work with to begin with, it needs to start getting some of these items quite a bit lower to generate profits. The good news is that there have been other individual markets (mostly in Europe) where Spotify has become profitable over time, and so there’s a precedent for profits following scale on a more local level.

Paid versus free streaming

The other interesting aspect of Spotify’s results is the split in its revenue sources, among which paid subscriptions and advertising account for over 99%. The split between those two has remained relatively constant over the last three years, with subscribers generating roughly 90% of revenue, and advertising the other 10%. That’s notable, because Spotify has over two times as many free subscribers as paid subscribers, but those paid subscribers generate nine times as much revenue. To look at it another way, the paid subscribers generate roughly 80 euros a year in revenue each, while the free subscribers generate just $3-4.

This has been a matter of some controversy within the music industry, but in that podcast episode I referred to earlier, we discussed this, and Ryan Wright’s take was that the existence of free streaming is actually important from a perspective of creating a funnel for future paid subscribers. Given that Spotify consistently has around one third paid subscribers, there’s some evidence that this funnel generates the desired results over time, but in the meantime there are many more subscribers generating vastly less revenue. It’s arguably critical both for Spotify and the broader music industry that this conversion of free subscribers continues, and the fact that Spotify reduced its price for family plans today to bring them in line with both Apple and Google is likely another attempt to boost this strategy.

Spotify dominates industry paid streaming revenue

The other interesting thing about this revenue is how it compares to overall industry revenue from both paid and ad-supported streaming. The IFPI is the global body representing the music industry, and its annual reports provide estimates of global revenue from streaming. This year, it broke out paid and ad-supported streaming specifically, and it’s worth comparing Spotify’s figures to those overall industry numbers. In order to compare these figures on a like-for-like basis, I’ve apportioned Spotify’s royalty costs on the same basis as its revenues between paid subscriptions and advertising, and compared those numbers with the IFPI’s industry revenue figures. On that basis, then, let’s look first at Spotify’s paid streaming revenue as a percentage of the IFPI’s industry revenue figure for this category:Spotify as percent of IFPI paid streaming revenueThe obvious conclusion is that Spotify’s growth is not just in line with industry growth but actually represents a significant gain in share of the total paid streaming market. On this basis, it’s clear that Spotify dominates overall paid streaming revenue for the industry (as it does subscriber numbers).

On free streaming, Spotify is the minority

The other interesting comparison is looking at Spotify’s ad-supported streaming revenue versus the number reported this year by the IFPI. The IFPI pegged ad-based streaming revenue at $634 million globally, which translates to 566 Euros based on today’s exchange rate. Spotify’s ad-based royalty payments were of the order of 164 million Euros, which translates to about a third of total industry revenues. In other words, its share of ad-supported streaming is a minority one, far from being dominant as it is in paid streaming. And that, of course, makes perfect sense when you consider YouTube’s role in ad-supported streaming. YouTube is likely far more dominant in usage on the free streaming side than Spotify is on the paid streaming side, but it pays out at a much lower rate, so its share of revenue is not as dramatic as its share of usage.

Useful context for an IPO

These numbers certainly make for interesting reading, and I’d love to get my hands on the full filing, because there are additional numbers in these documents which would be great fodder for additional analysis. However, even just with what I was able to glean from secondary sources, there’s plenty here to put Spotify’s rumored IPO plans in context. The company is growing fast, and dominates the paid streaming market. For investors looking to buy into the streaming trend, this looks like a great bet. Of course, the downside is that the company has yet to generate profits on a global basis, and it doesn’t look any closer to that milestone this year than last year. That’s something investors will want to look at very closely if and when Spotify does file to go public.

Apple Music’s other financial advantage

This is something I’ve though about quite a bit, and even wrote about briefly as part of a much longer piece ahead of the launch of Apple Music, but I feel like no-one is really talking about. But it’s potentially quite significant for the economics of Apple Music, and especially the per-stream payout rate Apple will end up passing on to labels and artists. Note: it’s already clear that Apple will have a higher per-stream payout simply based on the fact that it’s a paid-only service, whereas Spotify and other large services mix paid and free users. But I’m talking about an additional impact on top of that.

Integration of owned music is the key

The big factor here is Apple’s integration of the music you already own and store in iTunes into the Apple Music experience and into a single app. I think that’s huge for usability, and that was the key point in that earlier piece, but I think it could also be quite significant for the economics of the service. Why is that? Well, with almost any other subscription streaming service, you as the user tend to start from scratch in terms of your existing library. Perhaps you hop back and forth between apps when you play the music you own versus the music you’re streaming, but I’d guess many people just stick to a single app and play all their music from there, even the stuff they may have purchased somewhere along the way, because it’s all available in the streaming service and it’s easier to play it there than switch apps. Services like Spotify will pay out to artists regardless of whether the user already owns the track somewhere else (unless the user has imported their owned music). But when the user’s owned music is also available in the app, Apple won’t have to pay out when the user plays that music.

What’s really hard to know here is the balance between owned versus streamed music the average user plays during the course of a typical month. I know my own usage is heavily skewed towards the music I own and am familiar with, along with a few tracks or albums I don’t own and stream instead. Perhaps others are different, but I’d guess almost all users would spend a significant amount of time playing music they already own. With other services, the provider still has to pay out on this music because it all looks the same, but with Apple Music there will be a clear line between the music the subscriber owns and the music he or she is streaming through the service (even if it’s presented together in the context of the app). If the amount that’s played from the library rather than streamed is significant, this could substantially reduce the number of plays for which payments need to be made.

A higher per-stream rate on Apple Music

At this point, it’s worth thinking about how the economics of streaming music work. Although we often see per-stream rates used in discussions of how much artists get paid through these services, the reality is that there are no set per-stream rates. Rather, these services share some proportion of their overall revenue from subscribers and/or advertisers with those labels and artists whose music their subscribers listen to. The total pot is divided up with labels and artists according to a standard formula, and I’ve pasted the graphic Spotify uses to illustrate this formula below:

Spotify-Royalty-Formula

Once you understand that it’s a matter of dividing up the total pot, it becomes very relevant how many songs are streamed and therefore get to share in that pot. If Apple Music has fewer songs streamed through the service (because a significant proportion are played instead from users’ libraries), that in turn could dramatically increase the per-stream payout for those artists whose music is streamed. That will likely disproportionately benefit new artists and music over older artists and albums, which could be particularly good for those discovered through the service.

Of course, over time, this advantage will be mitigated as the balance between owned and streamed music shifts towards streamed music, as people will likely buy far less music (if any) going forward. But as the first Apple Music subscribers get past their trial periods and Apple starts paying out on its long-term formula, this could result in significantly higher payments per stream than other services. Add this to the existing advantage Apple has over competing services because of the paid-only nature of the service. Over time, that could have a really interesting impact on artists’ willingness to continue to work with other services. If, as an artist, you’re getting paid several times more on Apple Music per stream than on Spotify, Rdio, or Deezer, would you eventually consider pulling at least some of your music from those other services?

Note: I’m making a fundamental assumption here, which is that Apple will only pay out on plays of music the user doesn’t play from his or her own library. That seems a reasonable assumption, but I haven’t confirmed it. I can’t see why Apple would pay out on that music (unless it’s played through iTunes Match, which shares 70% with artists too), but it’s a remote possibility that it will, in which case the argument falls apart. 

Thoughts on Neflix’s Q1 2015 earnings

I’m kicking off the Q1 2015 earnings season (past earnings posts here) with a post on Netflix, just as I did last quarter. I’ll also be doing an updated deck for subscribers to the Jackdaw Research Quarterly Decks service. Having done a pretty broad run-down last time around, I’m going to focus on three things this time around:

  • Subscriber growth, especially in the domestic streaming business
  • Profitability of the US streaming business
  • Profitability of the other two businesses.

Subscriber growth in the US becomes ever more cyclical

As I said last time around, subscriber growth in the US is likely to slow down over time as the service reaches later adopters and much of the lower-hanging fruit is already harvested. However, Netflix had a really good quarter for net additions in Q1, and year on year additions were flattish compared to last quarter rather than down dramatically too:

Quarterly net addsYear on year sub growthSo what happened? Was I wrong about the long-term trend? Actually, no. What’s happening is that Netflix’s domestic subscriber growth in particular is becoming increasingly cyclical, driven heavily by new series launches in Q1 and lower in every other quarter. This is easier to see in this quarterly chart showing just US streaming subscriber growth: Continue reading