The Narrative Of Spotify’s Filing Is That The Best Is Yet To Come

Spotify just filed its F1 for its DPO. The most anticipated business event in the recorded music industry since, well…as long as most can remember, is one big step closer. The filing is a treasure trove of data and metrics, and while there won’t be too many surprises to anyone who follows the company closely, there are none the less a lot of very interesting findings and themes. The full filing can be found here. Here are some of the key points of interest:

  • Most of the numbers are heading in the right direction: MAUs, subscribers, hours spent etc are all going up while churn and cost ratios are heading down. Premium ARPU was an exception, declining: 2015 – €7.06 / 2016 – €6.00 / 2017 €5.24, which reflects pricing promotions. But Spotify was never going to have fixed every aspect of its business in time for its listing, that was never the point. What Spotify needs to convince potential shareholders is that it is heading in the right direction. The narrative that emerges here is of a company that has helped create an entire marketplace, that has made great ground so far and that is poised for even bigger and better things. That narrative and the clear momentum should be enough to see Spotify through. As I’ve previously noted, investor demand currently exceeds supply. If you are a big institutional investor wanting to get into music, there are few options. Pandora aside, this is pretty much the only big music tech stock in town. As long as Spotify can keep these metrics heading in the right direction, it should have a much smoother first few quarters than Snap Inc did, even though a profit is unlikely to materialise in that time.
  • Spotify is baring its metrics soul: Spotify has put a lot of metrics on the line, setting the bar for future SEC filings. While competitor streaming services will be busy plugging the numbers into Excel so they can compare with their own, the rest of the marketplace now has a much clearer sense of what running a streaming service entails. One really encouraging development is Spotify’s introduction of Daily Active User (DAU) metrics. As we have long argued at MIDiA, monthly numbers are an anachronism in the digital era, a measure of reach not engagement. So, Spotify is to be applauded for being the first major streaming service to start showing true engagement metrics.
  • Users and engagement are lifting: Spotify had 159 million MAUs in 2017, with 71 million paid and 92 million ad supported. Europe was the biggest region (58 million total MAUs) followed by North America (52 million), Latin America (33 million) and Rest of Word (17 million). The latter two are the fastest growing regions. Meanwhile, 44% of MAUs are DAUs, up from 37% in Q1 2015, which shows that users are becoming more engaged, though the shape of the curve (see chart below) shows that when swathes of new users are on-boarded, engagement can be dented. Consumption is also growing (a sign of both user growth and increased engagement): quarterly content hours went from 17.4 billion in 2015 to 40.3 billion in 2017. There are some oddities too. For example, ad supported MAUs actually declined in Q2 2017 by 1 million on Q1 2017 and in Q4 2017 only increased by 1 million on the previous quarter to reach 92 million.
  • The future of radio: Spotify puts a big focus on spoken word content and podcasts in the filing, as it does on advertiser products. It also lists radio companies first and subscription companies second as its key competitors. Meanwhile ad supported flicked into generating a gross profit in 2017 (ad supported went from -12% gross margin in 2016 to 10% in 2017. Premium gross margin up from 16% to 22% over same period.) As MIDiA predicted last year, free is going to be a big focus of Spotify in 2018 and beyond. The first chapter of Spotify’s story was about becoming the future of retail. The next will be about becoming the future of radio. And the increased focus on spoken word is not only about stealing radio’s clothes, it is about creating higher margin content than music. None of this is to say that Spotify will necessarily execute well, but this is the strategy nonetheless.
  • Spotify is still losing money but is trending in the right direction: Spotify’s cost of revenue in 2017 was €3,241 million against revenue of €4,090 with an operating loss of €378 million. However, losses are not growing much (€336 in 2016) and financing its debt added a whopping €974 million in 2017, from €336 million in 2016. Part of the purpose of the DPO is to ensure debt holders, investors and of course founders and employees get to see a return of their respective investments in money and blood, sweat and tears. Once that is done, financing costs will normalize. Also, Spotify’s new label licensing deals are kicking in, with costs of revenue as a share of premium revenue falling from 84% in 2016 to 78% in 2017. Spotify is not yet profitable but it is getting its house in order.

All in all, there is enough in this filing to both convince potential investors to make the bet while also providing enough fodder for critics to throw doubt on the commercial sustainability of streaming. Spotify’s structural challenge is that none of the other big streaming services have to worry about turning a profit. In fact, it is in their collective interest to keep market costs high to make it harder for their number one competitor to prosper.

But in the realms of what Spotify can impact itself, the overriding trend in this filing is that Spotify is well and truly on the right track. For now, and the next 9 months or so, Spotify will likely remain the darling of the sector. But after that, investors will start wanting a lot more if they are going to keep holding the stock. Spotify is promising that the best days are yet to come. Now it needs to deliver.

spotify f1 a

spotify f1 b


Radio Is Streaming’s Next Frontier

This week MIDiA held its latest quarterly research and networking event at Gibson Brands Showrooms in the heart of London’s West End. The event was heavily over-subscribed and was a great success (there are some photos at the bottom of this post).

The event combined a presentation from Pete Downton, deputy CEO of our event sponsor 7digital, a keynote from myself and a panel of leading industry experts. Here are a few highlights of my presentation.

radio blog slide

Streaming music has got where it has today largely by being the future of retail and replacing the download model, which in turn replaced the CD model (though vestiges of both remain). That premium model will continue to be the beating heart of streaming revenues for the foreseeable future but will not be enough on its own. The next big opportunity for streaming is to become the future of radio, which incidentally is around double the size of the recorded music market. In doing so, it will be a classic case of disruptive insurgents stealing market share from long-standing incumbents.

The opportunity for streaming is to build ad revenue around the younger audiences that are simply not engaging with traditional radio in the way that previous generations of young music fans once did. As the chart above shows, radio’s audience is aging and has an almost mirror opposite demographic profile to streaming. What is more, radio’s audience is declining by around one percentage point each quarter. It might not sound like much, but you normally do not measure change in terms of consistent quarterly trends. Instead there is normally quarterly fluctuation. So, this is nothing short of a major decline.

However, what is interesting is that free streaming is not growing by the same rate radio is declining. Instead, what is happening is that radio and streaming audiences are co-existing, with many that have spent a long time doing both eventually shifting all of their listening to streaming. Added to this, older consumers tend to embrace change more slowly than younger audiences. So, radio’s older listener base effectively acts as a disruption buffer.

What all this means is that radio is facing an existential threat like no other but it has some time to get its house in order, to identify how it can meld the best of the radio model with streaming experiences to start its fight back. And make no mistake, radio has so many unique assets that streaming does not (local content, talk, news, sports, weather, travel, brand personality etc.) and Apple’s underwhelming success with Beats 1 shows that hiring a bunch of radio people and launching a station does not guarantee success. Nonetheless, streaming services will get there. And Spotify’s recently launched Pandora-clone in Australia indicates just how serious the radio frontier is to streaming.

For more (a lot more!) data and analysis on how radio and streaming are facing up against each other, check out our new report Radio – Streaming’s Next Frontier: How Streaming Will Disrupt Radio Like It Did Retail which can be purchased directly from our report store here and is also available immediately to MIDiA clients as part of our research subscription service.

MIDiA Radio Event 1MIDiA Radio Event 2

Yonder And Streaming’s Less Travelled Path

Back in 2012, a music service that had raised $174 million in funding closed without yet having launched to consumers. That service was Beyond Oblivion, a company that intended to transform the music market with music bundled into handsets and phone packages at no extra cost to consumers. Five and half years later, Beyond Oblivion’s founder is finally seeing his latest iteration of the bundled music service model gain traction. Yonder, his new(ish) company, has started off 2018 with a million monthly active users (MAUs) under its belt, with the majority of that growth coming in the fourth quarter of 2017. Yet Yonder is not on many people’s radar, in large part because it is building its business in markets that are off streaming’s beaten track.

yonder graphic

Yonder’s main market is Bangladesh, which makes up just over half of its MAU base, followed by Indonesia and Sri Lanka. It even has tens of thousands of users in Nepal and the Maldives and plans to roll out to markets such as Myanmar, Cambodia, Iraq and Ghana in 2018. These are not markets famed as booming digital music markets, and they’re certainly not priority markets for any of the top streaming services. So, in many respects Yonder is competing around, rather than with the likes of Spotify.

Low ARPU markets

But there is more to it than just that. These are markets with mostly large populations and very low GDP per capita and mobile ARPU. In many of these territories mobile ARPU is significantly lower than the cost of a western streaming subscription. For example, total mobile ARPU in Bangladesh is around $4 a month. This makes fitting the economics of a streaming music bundle into a tariff challenging in the extreme. The standard wholesale tariffs record labels provide streaming services in these regions struggle to fit these wafer thin margins. So, making music bundles work needs a very specific and localized approach. The same principle applies to localization, with music programming requiring a much higher degree of local specialization than many other markets.

More than one way to skin a cat

2018 will likely see a slowdown in music subscriber growth in many western markets. In the meantime, majority of the 9.99 price points will be addressed. Ad supported and discounting will be key to sustaining growth in these markets, but the scale of opportunity for digital music lies in emerging markets. 2017 was the year we really started to see Latin American markets begin to make their mark, while China established itself as a major contributor to subscribers, if not revenue. Services like Yonder are important for the music business, not just because they address new markets but also because they represent another approach. The 9.99 AYCE model will remain the core opportunity, but sticking too tightly to it will limit the scope of the wider market.

Yonder’s model is not without challenges – not least the concept of making premium music feel like it’s free to its users – but it represents one of what should hopefully become a wider selection of alternative paths to making streaming pay.

Join Us At ‘Radio Is Streaming’s Next Frontier’

I’m very pleased to announce that MIDiA is hosting a special industry event on Wednesday 7th February at Gibson Brands in central London, in partnership with 7digital. The event ‘Radio Is Streaming’s Next Frontier’ is going to explore how in 2018, streaming music is going to start impacting radio just like it has spent the last few years replacing downloads. Streaming spent the first phase of its life being the future of retail, it will spend the next phase becoming the future of radio.

In this free-to-attend event we will present some of our latest research, including exclusive data, ranging from big picture trends through to tactical data, such as exactly how much each streaming service is affecting each radio station.

In addition to my research presentation there will be panel discussion from industry experts:

Is Streaming and Radio a Zero Sum Game?

Moderator: Zach Fuller


  • Jeff Smith: head of Music, Radio2 and Radio6
  • Pete Downton: deputy CEO, 7digital
  • Chris Baughen: VP Content and Formats, Deezer

After all this there will be drinks and networking. The event was publicised to MIDiA clients and newsletter subscribers first so there are only a few places left. So, RSVP your slot here now!

Hope to see you there, and watch out for a sneak peak of some of the research soon.


Free Report: Lyrics Take Centre Stage In Streaming Music

We are pleased to announce the publication of a brand new, totally free, Streaming Music report. In this report, we present the findings of an exclusive consumer survey fielded in November 2017 to consumers in the US, UK and Germany, deep diving into streaming behaviours and the growing role that lyrics is taking. The report download link can be found at the bottom of this post.

The report includes data on:

  • Overall music consumption and streaming behaviours
  • Weekly Active User (WAU) penetration of all key streaming music apps
  • Tenure splits of streaming users by streaming service
  • Consumer attitudes towards lyrics
  • Lyric users by tenure length of individual streaming services
  • The relationship between lyrics users and streaming loyalty
  • Key drivers for using lyrics, with gender splits

Here is an overview of some of the findings of the report that we wrote on behalf of LyricFind.


Streaming music has put the audience in control, letting music fans choose what, when and where they listen. One of the most dramatic changes that streaming has enabled is the expansion of music from a lean-back, linear experience into something far more engaging and interactive. Now fans lean forward to choose the songs they want, build playlists, comment and share. Lyrics are centre stage in this shift, transforming from static-print-hidden-away-inside-album-sleeve notes, to a dynamic extension of the music itself. Lyrics permeate the streaming music ecosystem, from websites, through YouTube and Vevo to the streaming services themselves.

Whereas lyrics in the analogue era used to be domain of music aficionados, in the streaming era they are a mainstream behaviour for audiences as diverse as they are widespread. The motivations are similarly varied, with the most cited being to know the words (81%) followed by being able to sing along (72%). Among streaming services users who are music subscribers, penetration of lyrics usage rises to 88%. What is more, lyrics have a strong link with music subscriber loyalty among 91% of all music subscribers that have been using lyrics for more than three years.

However, many lyrics users want more out of their lyrics experiences, with 56% of subscribers wanting lyrics to be in time with songs. Younger users, in particular, are raising their expectations, with 16-24 year olds the most likely to want new lyrics features.

Streaming is transforming music consumption across the board

Music consumption is in the midst of a transition period, with streaming rapidly ascending to become the dominant format. As with any transition, the old world coexists with the new, due to old habits dying hard and older groups of consumers changing behaviours slower. Thus, we see radio (66%) and free streaming (43%) as the two dominant forms of music consumption. Crucially, a strong overlap exists between the two: 72% of streamers listen to radio and 47% of radio audiences stream music. This indicates: a) that the transition will pick up pace, as nearly half of radio listeners are already swapping out some of their radio listening time for streaming; and b) that there currently remains enough that is different between radio and streaming for the two to coexist. The biggest takeaway though, is that streaming has a massive amount of growth potential ahead of it.

Lyrics are an integral part of the streaming experience

Lyrics are at the centre of the streaming music experiences: 79% of all music streamers use lyrics, rising to a comprehensive 88% of music subscribers. Wanting to know the words to songs is the main driver, with 65% of music subscribers stating this as their reason for using lyrics. Next, 55% of subscribers and 51% of free streamers said they wanted to be able to sing along with their favourite song. More social activities like singing with friends and karaoke score relatively lowly, indicating that lyrics are a very personal and integral part of how music fans interact with music.

Lyrics have a clear correlation with music subscriber tenure and with churn. The longer that consumers have been music subscribers, the more likely they are to use lyrics, while consumers that have cancelled their subscriptions are much less likely to use lyrics. Across Deezer, Spotify and Google Play, an average of 98% of subscribers with three plus years tenure use lyrics. This contrasts with lyrics penetration among churned out subscribers, with an average of just 60% across the same three streaming services. The importance of lyrics features is further underscored by the fact that 55% of streaming lyrics users say they are more likely to pay for a streaming service that ‘has great lyrics features’. For music subscribers overall, the rate is 43%, rising to 48% of Deezer users and 52% of Tidal users.

LyricFind - cover

Lyrics Take Centre Stage In Streaming – LyricFind – Report

MIDiA Research Predictions 2018: Post-Peak Economics

With 2017 drawing to a close and 2018 on the horizon, it is time for MIDiA’s 2018 predictions.

But first, on how we did last year, our 2017 predictions had a 94% success rate. See bottom of this post for a run down.


  • Post-catalogue – pressing reset on the recorded music business model: Revenues from catalogue sales have long underpinned the major record label model, representing the growth fund with which labels invested in future talent, often at a loss. Streaming consumption is changing this and we’ll see the first effects of lower catalogue in 2018. Smaller artist advances from bigger labels will follow.
  • Spotify will need new metrics: Up until now Spotify has been able to choose what metrics to report and pretty much when (annual financial reports aside). Once public, increased investor scrutiny on will see it focus on new metrics (APRU, Life Time Value etc) and concentrate more heavily on its free user numbers. 2018 will be the year that free streaming takes centre stage – watch out radio.
  • Apple will launch an Apple Music bundle for Home Pod: We’ve been burnt before predicting Apple Music hardware bundles, but Amazon has set the precedent and we think a $3.99 Home Pod Apple Music subscription (available annually) is on the cards. (Though we’re prepared to be burnt once again on this prediction!) 


  • Savvy switchers – SVOD’s Achilles’ heel: Churn will become a big deal for leading video subscription services in 2018, with savvy users switching tactically to get access to the new shows they want. Of course, Netflix and co don’t report churn so the indicators will be slowing growth in many markets.
  • Subscriptions lose their stranglehold on streaming: 2018 will see the rise of new streaming offerings from traditional TV companies and new entrants that will deliver free-to-view, often ad-supported, on-demand streaming TV.


  • Beyond the peak: We are nearing peak in the attention economy. 2018 will be the year casualties start to mount, as audience attention becomes a scarce commodity. Smart players will tap into ‘kinetic capital’ – the value users give to experiences that involve their context and location.
  • The rise of the new gate keepers part II: In 2018 Amazon and Facebook will pursue ever more ambitious strategies aimed at making them the leading next generation media companies, the conduits for the digital economy.


  • The rise of the unaffiliated eSports: eSports leagues emulate the structure of traditional sports, but they may have missed the point. In 2018, we’ll see more eSports fans actually seeking games competition elsewhere, driving a surge in unaffiliated eSports.
  • Mobile games are the canary in the coal mine for peak attention: Mobile games will be the first big losers as we approach peak in the attention economy – there simply aren’t enough free hours left in the day. Mobile gaming activity is declining as mainstream consumers, who became mobile gamers to fill dead time, now have plenty of digital options that more closely match their needs. All media companies need to learn from mobile games’ experience.


  • The fall of tech major ROI: Growth will come less cheaply for the tech majors (Alphabet, Apple, Amazon, Facebook) in 2018. They will have to overspend to maintain revenue momentum so margins will be hit.
  • Regulation catches up with the tech majors: Each of the tech majors is a monopoly or monopsony in their respective markets, staying one step ahead of regulation but this will change. The EU’s forced unbundling of Windows Media Player in the early 2000s triggered the end of Microsoft’s digital dominance. 2018 could see the start of a Microsoft moment for at least one of the tech majors. 

2017 Predictions

For the record, here are some of our correct 2017 predictions:

  • Digital will finally account for more than 50% of revenue
  • Spotify will still be the leading subscription service
  • eSports to reach $1 billion
  • Streaming holdouts will trickle not flood
  • AR will have hype but not a killer device.
  • VR players will double down on content spend
  • Google doubles down on its hardware ecosystem plays
  • 2017 will not be the year of Peak TV
  • Original video content to arrive on messaging apps

Here are some that we got wrong or were inconclusive:

  • Tidal finally sells ($300 million stake from Softbank was a partial sale – full sale likely in 2018)
  • Apple will launch an Apple Music iPhone – didn’t happen but the Home Pod may be the bundled music device in 2018 (see below)
  • Spotify will be disrupted – it actually went from strength to strength with no meaningful new competitor, yet

Shazam Is Apple’s Echo Nest

apple music shazam midia

Shazam finally found a buyer: Apple. Ever since its affiliate sales revenue model crumbled with the onset of streaming (there’s no business in an affiliate fee on a $0.01 stream), Shazam has been trying to find a new business model. It doubled down on providing tools for TV advertisers but never got enough traction for that to be a true pivot. Shazam’s problem has always been that it was a feature rather than a product – as so many VC funded tech companies are. The fact that it sold for $400 million – just 2.8 times its total investment ($143 million) and well below its previous pre-money valuation of $1 billion, illustrates how much value has seeped out of Shazam’s business. The Apple acquisition though, is one of the few ways that Shazam’s ‘hidden’ value can be realised.

Cool tech without a business model

Shazam was a digital music pioneer. I remember getting a demo from one of the founders back in the early 2000s, and I was blown away by just how well the tech worked. However, quality of tech was never Shazam’s problem, and once the app economy appeared it also had a very clear and compelling consumer use case. Despite competition from challengers in more recent years – especially Soundhound, which has also been compelled to pivot but may now decide to double back down on its core competences – Shazam continued to be the standout leader in music recognition. The irony is that its use case is stronger now than it was back in the download era because people are listening to a wider array of music than ever before. The problem was a lack of revenue model.

Shazam tried to position itself as a tastemaker, with its charts becoming useful heat indicators for radio stations and streaming companies. Labels soon learned to game the system with ‘Shazam parties’ but even without that challenge, this still did little to help Shazam build a business model. Apple however, saw beyond the music recognition and Shazam now gives Apple a music recognition engine. Shazam is Apple’s answer to Spotify’s Echo Nest.

Apple Music needs growth and engagement

Apple, which recently passed 30 million subscribers, continues to lag behind Spotify’s growth. Apple Music is adding around half a million new subscribers a month, while Spotify was adding close to two million a month up until it announced 60 million subscribers in July. The fact Spotify hasn’t announced since then may point to slowing growth, but my money is on a big number being announced in the next five weeks.

Apple’s weekly active user (WAU) penetration is far behind Spotify’s, indicating that Apple needs to do a better job of engaging its users. Better playlists, recommendations and algorithm driven curation all help Spotify stay ahead of the curve. Now, Apple will be hoping that Shazam will provide it with the tools to start playing catch up. And that’s not even mentioning the user acquisition potential Shazam could have when it switches to exclusively pointing to Apple Music. Game on.

Is Hip Hop Keeping Female Artists Out Of The Charts?

There is a gender divide in the upper echelons of popular music. In the US, Taylor Swift is the first woman to top the Hot 100 in 2017, with hits by women accounting for just 14% of all top 10 hits throughout the course of the year. But this is not just a US thing; it is a global trend, with female artists in the distinct minority in streaming too. There are various contributing factors, including unintended consequences of label A&R strategy and streaming service curation techniques.females in spotify

Looking at the global top tracks on Spotify, just 18% of the tracks from the top 30 streams are from solo female artists. In terms of chart positions the male dominance is even more pronounced with female artists making up 13%, or just four, of the top 30 slots.

So how did we get here? Part of the reason is that big female acts like Beyoncé, Rihanna and Katy Perry are out of cycle, but there are other factors too:

  • Hip hop A+R strategy: Hip-hop has replaced EDM as the record labels’ second favoured genre (after pop). When streaming exclusives were a thing, Apple Music and Tidal were fighting it out over hip hop and urban exclusives such as Frank Ocean and Beyoncé. Meanwhile, Drake was the most streamed artist of 2016. In 2014 and 2015, labels were falling over themselves to sign EDM artists. In 2016 and 2017, hip hop and urban artists have become the sought-after commodity. EDM is a male dominated genre but hip hop is even more so. Thus of the 11 hip hop artists in the top 30 most streamed tracks on Spotify globally for the week ending 8/11/17, all of them are male. You have to go down to position 33 to find the first and only solo female hip hop artist (Cardi B) in the top 50 streamed tracks.
  • Streaming algorithms: Playlists are at the heart of streaming consumption and their relevance is determined by a combination of user-data learning and human curation. As a result, a vicious circle emerges, whereby playlists end up full of the music labels are pushing, and because people tend not to skip that much when listening, the user data suggests that this is the music they want to listen to. To quote Paul Weller’s lyrics: the public wants what the public gets. With hip hop now de rigueur, it essentially self-accelerates on streaming. And with most of the big artists being male, females end up becoming side lined. It is a classic case of the law of unintended consequences.

In truth, many genres have a strong male bias. Rock, now a shadow of its former self (just two artists in the top 30 Spotify tracks are rock) is similarly male dominated. Pop and vocal music have long been the genres where female artists have done best. However, right now, the hip hop-defined picture of popular music is particularly skewed. When you get further down the tail, things even out a little. Within the top 50, female artists generate 23% of streams, and the full year numbers will be better balanced when seasonal skews, such as big female artist release cycles, are balanced out. Nonetheless, the inescapable takeaway is that the combination of label A+R strategy and streaming curation have, unintentionally, created a distorted picture of the top of the music pile.

Pandora’s Loss Is Sirius XM’s Gain

Pandora is in trouble, as explained by the consistently excellent Tim Ingham at Music Business Worldwide, after losing a billion dollars over the last four years and monthly active users (MAU) fell to 73.7 million – its lowest point since Q1 2014. Regular readers will know that I’m a long-time advocate of Pandora’s model. Indeed, Pandora’s model is the future of radio. However, it now appears that Pandora may not be the future of Pandora’s model. In fact, with Liberty Media subsidiary Sirius XM waiting in the wings for Pandora’s market cap to fall even lower than its current $1.4 billion (down from $8 billion in Q1 2014), Pandora might not even be the future of Pandora. In fact, Pandora’s struggles could be Sirius XM’s gain, exactly when it needs the help.

Pandora’s three most important metrics have long been:

  1. MAU
  2. Revenue
  3. The share of total radio listening it accounts for

All three are intertwined, but Pandora has managed to sustain strong growth in numbers two and three because it got better at increasing engagement and driving ad revenue from a largely flat MAU base. However, Pandora was only ever going to be able to squeeze so much revenue out of a flat user base. So, it is no surprise that ad revenue for the nine months to September 2017 was up a paltry 2.4% at $777.3m, compared to the same period in 2016 (figure). Pandora’s problem is not monetization. Indeed, it is better at monetizing ad supported streaming than any other player on the planet, having invested heavily in ad sales infrastructure and continuing to innovate ad formats. But even the shiniest car will eventually grind to halt if it has a gaping hole in its fuel tank. And make no mistake, Pandora has a gaping hole.

Spotify Stole Pandora’s Clothes

Long before Spotify was changing the music business, Pandora was virtually single-handedly creating the US streaming market – though subscription service Napster (then Rhapsody) was also making a small contribution. For the best part of a decade Pandora had almost all of the market to itself, but it is now buckling under the impact of on-demand streaming. Pandora was meant to be different to Spotify, and it was, until Spotify started stealing Pandora’s clothes. Pandora grew its user base by delivering a lean back, but personalized listening experience. Radio on its users’ terms. Spotify soon recognized the value of lean back listening, bringing in a vast selection of curated playlists, directly and via partners. Beats Music followed suit and soon became the foundation for Apple Music’s curated streaming proposition.

Pandora’s Reach Metrics Obscure The Real Story

Pandora’s own key metrics have been part of the problem. It fell into the same trap that traditional radio broadcasters did, of convincing itself that its reach metrics were a genuine indicator of its success. But reach means nothing in the digital era. Engagement is everything. MAU is a meaningless metric in today’s always on world. If you have an app on your phone that you only use once a month, you’d hardly consider that active usage. Active usage is measured at the very least in weekly active user (WAU) terms. That’s why at MIDiA we track all digital media apps using this measure to reveal just how active user bases really are.

midia pandora sirius xm

On this basis, Pandora has jut 22% WAU penetration in Q3 2017, representing around 57 million users, or 77% of its MAU base. That ‘missing’ 17 million users will be the ones that Pandora will lose next over the coming 12 months. Yet, its WAU base is at risk too. 26% of Pandora’s WAUs – its most engaged users – also use Spotify. Although Pandora has done an admirable job of building its own subscription business – reaching 5.1 million subscribers in Q3 put it at a credible sixth in the global subscriber rankings, it is looking like it’s too little too late. Furthermore, dumping its founder Tim Westergren robbed Pandora of a genuine visionary just when the company needed him most.

Pandora Will Enhance Sirius XM’s User Base

Pandora’s loss will be Sirius XM’s gain. Sirius XM has been feeling the pressure from Spotify and co, just like Pandora, but it has also experienced competitive pressure from Pandora. Sirius XM is another of radio’s potential futures, but it has faced growing pressure from Pandora and also other streaming services. The growing adoption of interactive dashboards in cars has been key (5% of US consumers now have one). Sirius XM’s WAU base fell from 11% of consumers in Q4 2016 to 8% in Q3 2017. That 30 decline is far more dramatic than Pandora’s 6% WAU decline over the same period. The 8% WAU penetration represents around 21 million users which means that its active user rate is even lower than Pandora’s at just 69%. Added to that, more of Sirius XM’s WAUs (30%) use Spotify. It also has a demographic time bomb ticking: just 8% of WAUs are aged under 35 while just 49% are female. This compares to 31% and 57% respectively for Pandora. Sirius XM’s aging user base is old and male. While Pandora’s is young(er) and female. This is Sirius XM’s opportunity.

In 2016, Sirius XM made an informal offer of $3.4 billion for Pandora. Today, it looks like an amazing deal for Pandora, but Pandora turned it down. Sirius though was not deterred and was able to get close to its goal by investing $480 million in a struggling Pandora in June 2017 and securing three board positions. Now all Sirius has to do is wait for Pandora’s stock to fall further and make its move – perhaps when the market cap gets closer to $750 million. When this happens, Sirius will get a major boost to its user base. More than that though, Sirius will significantly enhance its audience profile. Sirius and Pandora’s user bases are so different in composition that they will slot together like jigsaw pieces. The challenge for Sirius will be how to integrate Pandora in terms of feature sets, user experience, business model and, of course, company organisations. That challenge could prove even bigger than Pandora’s attempted turn around.

The data in this blog post is taken from MIDiA’s forthcoming report: Radio – Streaming’s Next Frontier: How Streaming Will Disrupt Radio Like It Did Retail 

If you are not yet a MIDiA client and would like to find out how to get access to this report email

Announcing MIDiA’s Streaming Services Market Shares Report

coverAs the streaming music market matures, the bar is continually raised for the quality of data required, both in terms of granularity and accuracy. At MIDiA we have worked hard to earn a reputation for high-quality, reliable datasets that go far beyond what is available elsewhere. This gives our clients a competitive edge. We are now taking this approach a major step forward with the launch of MIDiA’s Streaming Services Market Shares report. This is our most comprehensive streaming dataset yet, and there is, quite simply, nothing else like it out there. Knowing the size of streaming revenues, or the global subscriber counts of music services is useful, but it isn’t enough. Nor even, is knowing country level streaming revenue figures. So, we built a global market shares model that breaks out subscription revenues (trade and retail), subscribers, and subscription market shares for more than 30 music services at country level, across 30 countries and regions. You want to know how much subscription revenue Spotify is generating in Canada? How many subscribers Apple Music has in Germany? How much subscription revenue QQ Music is generating China? This is the report for you. Here are some highlights:

promo slide2 

  • At the end of 2016 there were 132.6 million music subscribers, up from 76.8 million in 2015
  • In Q4 2016 Spotify’s subscriber market share was 35% and it had $2,766 million in retail revenue
  • Apple Music was second with 21 million subscribers at the end of 2016, a 15.6% market share and it had $912 million in retail revenue
  • In 2016 Apple was the largest driver of digital music revenue across Apple Music and iTunes
  • The US is the largest music subscription market, which Spotify leads with 38% subscriber market share
  • The UK is Europe’s largest streaming market, which Spotify also leads
  • China’s subscriber base is the second largest globally, but it ranks just 13th in revenue terms
  • Japan is the world’s third largest subscription market, in which Amazon has the largest subscriber market share
  • Brazil is Latin America’s largest music subscription market

The report contains 23 pages and 13 charts with full country detail as well as audience engagement metrics. The dataset includes four worksheets and a comprehensive methodology statement.

Streaming Services Market Shares is available right now to MIDiA premium subscribers. If you would like to learn more about how to access MIDiA’s analysis and data, email

The report and data is also available as a standalone purchase on MIDiA’s report store as part of our ‘Streaming Music Metrics Bundle’. This bundle additionally includes MIDiA’s ‘State of The Streaming Nation 2.1’. This is our mid-year 2017 update to the exhaustive assessment of the streaming music market first published in May. It includes data on revenue, forecasts, consumer attitudes and behaviour, YouTube, app usage and audience trends.

Examples of country graphics (data labels removed in this preview)

promo slide7