Pandora Buys Rdio To Become A Global Streaming Powerhouse


pandora rdioPandora today announced that it was acquiring the assets of now failed subscription service Rdio.  While the whispers about Rdio’s future had been building for some time, the deal is more interesting for what it says about Pandora’s plans than what it says about the state of the subscription business.


Rdio Battled Bravely And Set Innovation Standards But Fell Short

For what Rdio lacked in subscriber numbers it made up for in innovation.  It continually set product and feature precedents that Spotify and others subsequently aped, and its $75 million dollar ad inventory deal with US radio giant Cumulus sets a business model blueprint that other streaming services will follow. But for all its efforts and extensive marketing efforts Rdio was simply not able to get to the same sort of level as Spotify’s 2nd tier competitors, let alone to seriously challenge Spotify itself.  The music subscription business is not a winner-takes-all market.  But it is one in which some degree of meaningful scale is required to trigger the telco partnerships and brand advertiser deals that are necessary to achieve sustainability.  Eventually a company transitions from ‘bright new hope with potential’ to an ‘also ran that isn’t ever going to make it’.  Once that imperceptible line of market perception has been crossed it is only a matter of time before the end comes.

Pandora Will Use Rdio’s Assets To Go Global

Crucially Pandora is not acquiring Rdio as a going concern but only its assets, which won’t include licenses (as they have to be renegotiated when a music service changes corporate hands).  What those assets represent, or at least the bits that matter to Pandora, are teams, product and tech, licensing know how and an international footprint.  That last bit is particularly pertinent.  Rdio’s 100 markets contrasts sharply with Pandora’s 3 (US, Australia and New Zealand).  Indeed Pandora CEO Brian McAndrews stated “We seek to be the definitive music source for music discovery and enjoyment globally”.  While 100 markets is probably a step too far for Pandora, expect a healthy selection of top tier and emerging markets to feature in Pandora’s roadmap.  And if you’re eager to identify which ones, just take a look at the bigger radio markets globally (Japan possibly excepted).

Pandora’s Success Is Built On Lean Back Not Lean Forward

Pandora’s success is firmly rooted in delivering a high quality, lean back experiences to largely mainstream audiences.  That’s how it reaches 78 million monthly listeners, more than a quarter of US adults.  That positioning has served Pandora well and made it one of the few success stories of digital music.  In fact, other than Beatport and Last.FM, it is one of the very few music start ups that had an exist that considered to be a true financial success. Crucial to that success has been the fact Pandora has operated under statutory licenses for semi-interactive radio, which leaves it with dramatically higher (potential) operating margins than on demand services.  Which begs the question, just why is Pandora getting into the subscription business?

This Is The Latest Part Of A Major Strategic Pivot

The answer is that it forms part of a much bigger, much bolder plan.  Pandora has spent the last couple of years quietly amassing the assets that will transform it into a music platform super power.  In 2015 it acquired music data company Next Big Sound (c.$50 million), then came ticketing company Ticketfly in October ($450 million) and now Rdio ($75 million).  The combined $0.6 billion is a truly sizeable investment in a streaming-centred business model by anyone’s standards.  It also accompanies a concerted and costly investment in Pandora’s regional ad sales teams across the US to compete on a level footing with traditional radio’s sales teams.  Couple all that with November announcements to become the exclusive streaming outlet for popular podcast series ‘Serial’ and the landmark direct deal with Sony/ATV Publishing and a picture of something truly ambitious starts to emerge.

Pandora was fortunate to be able to IPO at a time when public offerings were still a highly viable option for digital start ups.  Spotify and Deezer (which just cancelled its IPO) will look on with no little jealousy at the power that a market capitalisation of nearly $3 billion gives you.  Now it is using this financial firepower to take the next step on its streaming journey.  Whatever that will prove to be, expect it to be a platform in its truest sense, rather than simply a streaming service with a few loosely attached ‘alternative revenue’ models, which is a mistake some of the subscription incumbents have made thus far.

Discovery Doesn’t Lead Anywhere Anymore, At Least Not To Sales

Pandora may aspire to be the definitive source of ‘music discovery’ but streaming discovery is becoming streaming consumption.  i.e. it is increasingly not leading to sales.  Live music sales is one alternative way to make money from ‘discovery’ but if ‘free music to sell tickets’ is Pandora’s end game then some difficult conversations with songwriters (who of course often don’t play live) will need to be had.

Pandora has just thrown its hat into the ring as a top tier player in the global streaming business.  By some measures you could say it is poised to become the biggest.  McAndrews left no room for doubt by stating “We plan to substantially broaden our subscription business.”  But in doing so Pandora will have to look itself in the mirror and ask itself “what am I now?”.




Why Streaming Doesn’t Really Matter For Adele

The outstanding success of Adele’s single ‘Hello’ has stoked up the already eager debate around whether Adele’s forthcoming ‘25’ album is going to be a success.  Indeed some are asking whether it is going to ‘save the industry’. One of the aspects that is getting a lot of attention is whether the album is going to be held back from some or all of the streaming services.  The parallels with Taylor Swift’s ‘1989’ are clear, especially because both Swift and Adele are strong album artists, which is an increasingly rare commodity these days. But the similarities do not go much further.  In fact the two artists have dramatically different audience profiles which is why streaming plays a very different role for Adele than it does for Swift.

Lapsed Music Buyers Were Key To the Success Of ‘21’

Adele’s ’21’ was a stand out success, selling 30 million copies globally.  Core to ‘21’s commercial success was that the album touched so many people and in doing so pulled lapsed and infrequent music buyers out of the woodwork.  The question is whether the feat can be repeated? In many respects it looks a tall ask.  We’re 4 years on since the launch of ‘21’ and the music world has changed.  Music sales revenue (downloads and CDs) have fallen by a quarter while streaming revenues have tripled.  And the problem with pulling lapsed and infrequent buyers out of the woodwork is that they have receded even further 4 years on.  In fact a chunk of them are gone for good as buyers.

buyer streamer overlap

But beneath the headline numbers the picture is more nuanced (see graphic).  Looking at mid-year 2015 consumer data from the US we can see that music buyers (i.e. CD buyers and download buyers) are still a largely distinct group from free streamers (excluding YouTube).  While this may seem counter intuitive it is in fact evidence of the twin speed music consumer landscape that is emerging.  This is why ‘Hello’ was both a streaming success (the 2nd fastest Vevo video to reach 100m views) and a sales success (the first ever song to sell a million downloads in one week in the US).  These are two largely distinct groups of consumers.

Streaming A Non-Issue?

As a reader of this blog you probably live much or most of your music life digitally, but for vast swathes of the population, including many music buyers, this is simply not the case.  Given that the mainstream audience was so key to ‘21’s success we can make a sensible assumption that many of these will also fall into the 27% of consumers that buy music but do not stream.  The implication is thus that being on streaming really is not that big of a deal for ‘25’ one way or the other.  Whereas Taylor Swift’s audience is young and streams avidly, Adele’s is not.  That is not to say there aren’t young Adele fans, of course there are, but they are a far smaller portion of Adele’s fan base than Swift’s.

60% of 16-24 year olds stream while just 20% buy CDs.  Compare that to 40-50 year olds where 34% stream and 43% buy CDs.  These are dramatically different audiences which require dramatically different strategies.  Audio streaming is unlikely to be a major factor either way for Adele, neither in terms of lost sales nor revenue.  Unless of course she ‘does a Jazy-Z‘ or ‘does a U2’ and takes a big fat cheque from Apple to appear exclusively on Apple Music.  But I’d like to think she’d like to think she’d have the confidence of earning sales the real way.

The Importance Of The Digitally Engaged Super Fan

What unites Swift and Adele is that they are both mass market album artists and as such are something of a historical anomaly.  Swift bucked the trend by making an album targeted at Digital Natives shift more than 8 million units.  Adele will likely also buck the trend.  But paradoxically, considering the above data, in some ways it will be a harder task for Adele.  Swift has a very tightly defined, super engaged fan base that identifies itself with her.  Adele’s fanbase is more amorphous and pragmatic.  You don’t get ‘Adelle-ettes’.  Swift was able to mobilise her fanbase into music buying action like a presidential candidate with a passionate grassroots following and big donors.  The importance of digitally engaged super fans is the secret sauce of success for digital era creators.  It is the exact same dynamic that ensured UK YouTuber Joe Sugg was able to leverage his fanbase to give his debut book ‘Codename Evie’ the biggest 1st week sales for graphic novel EVER in the UK this year.

If Adele and her team do pull off a sales success with ‘25’ they will owe a debt of gratitude to that 27% of consumers.  While the odds are against it being quite as big as ‘21’ (simply because the market is smaller) it still has every chance of being a milestone event that will out perform everything else.  But do not mistake that for this being ‘Adele saves the music industry’.  Album sales are declining.  Success from Taylor Swift and Adele are (welcome) throwbacks and they are most certainly not a glimpse into the future.

Apple Music By The Numbers

Back in August when Apple announced it had hit 11 million subscribers I predicted that would result in around 6 million paying subscribers.  Yesterday Tim Cook announced that Apple Music now has 6.5 million paying subscribers, which translates into a 59% conversion rate.  Or at least 59% of trialists paid for at least one month.  As I wrote back in August, Apple will lose a share of those subscribers who will cancel after one payment (i.e. the ones who’d forgotten to cancel their payment details).  Somewhere north of 1.5 million of those subscribers will likely not make it through to a second month’s payment.  Which would leave around 5 million of those as long term subscribers.

The Acquisition Funnel Needs Widening

Cook also stated that the total number of users is 15 million which means that there are 8.5 million active trialists. Given that all the 11 million trialists reported 6 weeks after launch are now either gone or are subscribers that means all of those are additional trialists which gives us a monthly trialist rate of under 3 million or a little under 100,000 a day. Which is way below the 315,000 a day Apple had during the first 6 weeks (which is to be expected) but also below the 175,000 rate I had conservatively predicted back in August.  So Apple’s funnel is not yet performing as strongly as expected.  Given that most of Apple’s advertising for Apple Music is branding focused at the moment, we could expect that rate to augment steadily over the coming year as that brand message beds in.  And it could lift significantly if Apple shifts focus to product centric marketing i.e. what it normally does. (The Apple Music ad campaign is rare for Apple in that it doesn’t involve any product imagery).

apple music infographic

10 Million Cumulative Subscribers By Year End

If Apple continues at the current rate it should get to around 10 million subscribers by year end, of which 6 million or so will be active (i.e. not churned).  Which is again below my August prediction of 8.7 million because the acquisition funnel isn’t delivering as anticipated.  In revenue terms that would deliver cumulative subscriber revenue of $220 million by the end of the year.  Apple has earned around $140 million in total so far, of which $100 has gone to rights owners.

And we shouldn’t understate the scale of Apple’s success so far, narrow funnel or not.  It took Spotify 4 and a half years to get to 6.5 million subscribers.  Granted, it was a very different world back then and much of that growth had come without the US and of course without the benefit of Apple’s integrated ecosystem.  But even those considerations accounted for, Apple has gone from zero to hero in a flash.  In August I stated ‘Apple is on track to be the number 2 streaming subscriptions provider after little more than 6 months in the game’ and that is exactly where they are now.

To Restate Or Not To Restate

Music Business Worldwide cites an insider source that Spotify is on the verge of announcing its own new numbers. It will be interesting to see the fine print of how those numbers are reported.  Spotify has seen an uptick in subscriber growth at the same time it introduced its $1 a month for 3 months promotion, which is effectively a paid extended trial.  Here’s the conundrum.  If those numbers are reported as subscribers then expect terrible churn (for subscriber numbers) but if they are reported as trialists then conversion rates will be great but total subscriber numbers will not.  Common sense would dictate Spotify reporting those numbers as subscribers (they are paying after all) but that means at some stage Spotify is going to have to restate its numbers or provide some additional guidance.  Which incidentally Apple will also eventually have to do if it reports it cumulative 10 million subscribers at year end / early 2016 rather than the active subscriber number of around 6 million.

Apple and Spotify are now locked in a metrics arms race.  Both will use every trick in their respective arsenals to make those numbers look as good as they possibly can.  Whatever the outcome of that particular little spat, today’s numbers show us that even below its best, Apple just ran the first lap of a 5,000 metre race as if it was a 100 metre sprint. Let’s see if Apple can run an entire Mo Farah race at the speed of an Usain Bolt sprint.

The Global Implications Of The BBCs Streaming Strategy

Yesterday the BBC’s Director General Tony Hall laid out a vision for the future of the BBC (for an excellent take on this see the blog post from MIDiA’s video analyst Tim Mulligan, and yes the name may look familiar, he’s my brother!).  The BBC has long played a crucial innovation role in the digital content economy but it has yet to carve out a convincing role for itself in online music.  It has built up a compelling YouTube content offering and it has pursued a streaming coexistence strategy with its innovative Playlister initiative but the bigger play has yet to be made.  That looks set to change, with the announcement that the BBC is planning to launch a ‘New Music Discovery Service’, which would make the 50,000 tracks broadcast by the BBC every month available to stream for a limited period.  The initiative is interesting in itself but its implications are more profound and could have global repercussions.

Radio Still Rules The Roost But The Streaming Fox Is At The Door

Radio is still by far the main way most people interact with music.  75% of consumers listen to music radio regularly compared to 39% that stream for free. Radio also remains the main way in which people discover new music and its DJs are still some of the most influential tastemakers on the planet cf Apple poaching Zane Lowe from the BBC’s Radio 1.  But things are undoubtedly changing.  Music radio penetration among 16-24 year olds falls to 65% while streaming rises to 54%.  In Sweden streaming has overtaken music radio among 16-24 year olds.  All of this without even considering YouTube which has overtaken radio for 16-24 year olds in markets as diverse as UK, US, Sweden, Germany and Mexico and is on the verge of doing so in France.  (All consumer data is from MIDiA Research).  Radio held its own throughout the digital revolution of the last 15 years but the cracks are now there for all to see.  Most radio broadcasters do not yet have the assets to properly navigate the digital transition.  In most markets there is no dedicated digital platform (the US and UK are two notable exceptions) so broadcasters rely increasingly on mobile streaming for engaging audiences digitally.  Which means they are one swipe of a finger away from a bewildering array of radio alternatives.  It is this dynamic that underpins the BBC’s approach to streaming.

The Tyranny Of Choice

Though streaming had been around long before Spotify (hello Rhapsody) the Swedish upstart simply made the model work.  It did so by fixing buffering and by giving consumers frictionless (i.e. not cost and easy to use) access to all the music in the world.  By fixing that problem Spotify inadvertently created a new problem: the Tyranny of Choice.  Consumers are paralysed by excessive choice.  The Tyranny of Choice is of course not solely Spotify’s fault but it was certainly a catalyst for it. With the traditional gatekeepers / curators (delete as appropriate according to your worldview) increasingly bypassed by data-driven discovery, mainstream music fans are left feeling utterly bewildered.

Consumers Don’t Get Curation

The BBC is keenly aware of its value as a curator and quite frankly thinks it can do a better job than pure play streaming services.  It is probably right.  But what it doesn’t yet know how to do is communicate and deliver that value outside of the framework of radio.  The problem with curation is most people don’t think they need it.  Just 5% of consumers state they want discovery and recommendation features from streaming services.  Yet these are in the main the very same consumers that listen to music radio, which of course is all about discovery and recommendation.  The difference is that it doesn’t feel like it.

Setting Curation Free

This the challenge for the BBC and all radio broadcasters: how can they take the essence of DJ led programming and translate that into the streaming environment.  Apple’s approach of simply taking programmed radio and building on demand streaming around it is one bold approach but it is just a first step. The BBC, and other publicly funded broadcasters, have the advantage of being able to take the long view, of planning for long term evolution rather than focusing on ‘flipping’ their start up or keeping shareholders happy each quarter.

The BBC is placing the bet that giving its curation the maximum ability to permeate and interact with the streaming marketplace will give it the best chance of delineating which models will work and how best to bottle up that curation magic dust.  It is also a bold move because if it follows its course this could see the BBC’s content, curation and editorial break free of the confines of the BBC.  Because if it works well enough out in the ‘streaming wild’ why would a user need to even visit a BBC property.  The BBC is setting curation free.  It is a strategy that gives a hat tip to BuzzFeed, a company with a stated intent to distribute content as widely as possible even if that ultimately means killing off the BuzzFeed website.  A quote from BuzzFeed’s CEO Jonah Peretti sums up the thinking perfectly: “Content might still be King but distribution is Queen, and she wears the trousers.”

So watch the BBC’s streaming endeavours closely because the outcomes will likely provide blueprints for thriving in the streaming era for media companies of all types and sizes right across the globe.

Why Niche Is The Next Streaming Frontier

If 2014 was the year of fear, uncertainty and doubt for streaming then 2015 is shaping up to be the year in which streaming starts to deliver.  In fact so far streaming has helped drive revenue growth in the first half of 2015 for markets as diverse as Italy, Spain and Japan as well as of course in the streaming Nordic heartlands of Sweden, Denmark and Norway.  All this despite an accompanying average decline in download revenue of 7%.  But as I have long said, there is only so far that 9.99 AYCE (All You Can Eat) subscriptions can go.  This value proposition and price point combination constrains appeal to the aficionados and the upper end of the mainstream.  Pricing will be key to unlocking new users (as Spotify’s focus on the $1 a month for 3 months promo shows). However some highly influential elements within major labels are more resistant to pricing innovation now than they were this time last year.  So don’t hold your breath for the long overdue pricing overhaul.  The other side of the 9.99 AYCE equation though is just as important, namely choice, or rather, less choice. In fact, done right, cut down, niche music offerings should be able to fix the pricing conundrum too.

Too Much Content Is No Value At All
catalogue anatomy

Most people are not interested in all the music in the world and most people are not interested in spending $9.99 (or the local market equivalent) a month for music.   All the music in the world is a compelling proposition for super fans, but it is both a daunting prospect and more than is required for casual fans.  In fact the supposed benefit becomes a problem, the excess of choice begets the Tyranny Of Choice.  Indeed, just 5% of streaming catalogues is regularly frequented.  Most of the rest is irrelevant for most consumers.

Cord Nevers Are A Music Industry Problem Too

Most music fans like one or more kinds of music most.  While super fans are happy to pay for the ability to get everything, mainstreamers are not.  This is exactly the dynamic we are seeing in the video space, with consumers increasingly turning to smaller, cheaper services such as Netflix and Amazon rather than paying through the nose for an excess of cable channels.   The TV industry calls these consumers cord cutters (i.e. those that cancelled their TV subscriptions) and cord nevers (i.e. those that never paid for cable).  Now the music industry is facing its own cord never challenge: consumers who have never taken up a music subscription and have no intention of doing so.  In the past they would have spent some money on downloads, now they’re just watching more music videos YouTube.  The music industry quite simply does not have a Netflix for its cord nevers to go to instead of the full priced subscription option.

The Case For Niche Playlist Services

But give those more casual music fans a music app just built around their tastes and for a fraction of the price and the equation changes from zero sum.  Imagine genre specific playlist apps for $3 or $4 month.  A dozen curated playlists, a handful of featured albums and a couple of radio stations, all just of your favourite style of music and all streamed into a dedicated app.  Not only does this proposition deliver clear value, it also gives the industry an opportunity to open up new users that have thus far not been swayed by the broader utility play of AYCE services.

Imagine a Country app, a Classic Rock app, a Hip-Hop app, a Metal app, an EDM app, a Jazz app…. Each of these would create clear appeal within the mainstream elements of genre fan bases.  And while there is some risk of cannibalizing $9.99 services, this should be small if they are 100% curated (i.e. no on demand element) because they would be unlikely to appeal to aficionados and the super-mainstream.  These niche music apps could be delivered by standalone curated playlist service providers like MusicQubed, white label providers like Medianet and Omnifone, or even by AYCE services like Spotify ‘doing-a-Facebook’ by spinning out standalone apps.

The Marketplace Needs Niche Services Right Now

Niche services are not however a nice-to-have, an optional extra for the industry.  They will be crucial to unlocking the scale end of the subscription market and they will be needed sooner rather than later. Organic subscription growth (i.e. not including the temporary adrenaline shot of Spotify’s limited time price promotions) is not growing fast enough.  Apple Music looks set to add a significant amount of new users before year-end but many of those will come at the direct expense of the incumbents.  All the while YouTube is leaving everyone else for dust: the amount of net new video streams (i.e. free YouTube views) in H1 2015 was more than double that of net new audio streams.

The 9.99 AYCE model still has a lot of life in it yet, but just as the mobile phone market has far more choice than high end devices, so the subscription market desperately needs the diversity that niche services would bring.

Why Profit Doesn’t Come Into It For Apple Music

Apple has only ever been in the music business in order to sell more devices.  Apple does not need to make money from music nor has it ever needed to.  That doesn’t stop it being a crucially important music industry partner (in fact Apple is still pretty much the single most important partner on a global basis).  Nor does that mean that Apple doesn’t care about music or that it doesn’t take its role in the marketplace seriously.  But Apple is not in this game to make money.  Apple routinely ran the iTunes Store at ‘an about break-even basis’ which is financial report code for ‘at a slight loss’.  (Or in fact probably at a big loss if half of the costs of the combined iTunes / iPod ads had been factored in.)  Now Apple is spending big again on marketing its music product, but this time the ads are only for Apple Music so costs can’t be attributed to other parts of the business. Why this all matters is because it shows us just how seriously Apple is taking Apple Music and also its appetite for running it a loss leader.

Why Doesn’t Apple Just Buy Spotify?

One of the recurring questions around Apple’s streaming strategy is ‘why doesn’t it just buy Spotify?’.  Besides the fact it had already acquired Beats Music as part of the much bigger Beats purchase, Apple is not in the business of running other companies’ services.  Apple runs Apple services. This is because Apple is first and foremost a hardware business and its software and services are an extension of this – part of the device value proposition.  If Apple was a software and services business it would build Mac OS, iLife, iWork etc for other platforms.  Apple even made music production software Logic Mac only after buying it from eMagic.  iTunes is one of the stand out exceptions for this strategy but it is a legacy of when iPod was a PC / Mac centric device, where not being on Windows would have stymied iPod growth. (There is of course talk of Apple Apple Music becoming available on Android but if it does so it will only be because Apple wants to win back iTunes customers from Spotify.)

A Tunnel Vision Commitment To User Experience

The hardware-first / Apple-only strategy means that when Apple does buy other services it usually either assimilates them wholesale (remember LaLa?) or it strips them down to the bare bones and rebuilds them entirely (Beats Music).  This is all because Apple needs to own the customer relationship and customer experience in its entirety.  Apple’s tunnel vision commitment to user experience is the ideology that underpins this entire approach.  Which is why Apple didn’t buy Spotify.

Apple Could Make Most Streaming Margin By Promoting Spotify

apple music margin calcs

In fact Apple could make a LOT more money if it simply decided to spend money marketing Spotify to iOS customers.

For argument’s sake let’s assume Spotify has somewhere in the region of 6 million US subscribers, that 60% of those are on iOS and that 60% of those iOS users pay via iTunes, Apple thus generates $8.4 million a month in subscription revenue from Spotify.  To generate the same amount of US subscription margin from Apple Music, Apple would need 16.9 million US Apple Music subscribers (assuming an operating margin of 5%).  In fact, in practice Apple will be in heavy negative margins with Apple Music due to its extensive marketing efforts.

So if Apple was in the business of music for making money it wouldn’t even buy Spotify, it would simply spend money marketing it to the Apple customer base.  But that has never been the Apple way and is patently unlikely to become the Apple way. Thus Apple will continue on its mission to own every ounce of the streaming subscriber’s user journey.  Unfortunately the rest of the marketplace has to try to figure out how to compete while at the same time vainly searching for a profit.

The Real Problem With Streaming

Much of the debate around the sustainability of streaming has understandably focused on artist and songwriter income and transparency.  It is a debate that I have contributed to frequently.  But the more fundamental structural issues are whether the business models are commercially sustainable and if they are, what the implications are.  Music consumption is inarguably moving towards access based models so the question is not whether streaming should happen or not, but how to make it work as well as it possibly can for all parties.  As unfair as it might seem, the baseline issues regarding creator income could go unchanged without streaming business models falling apart.  But, as I will explain, if broader commercial sustainability issues are not fixed then many streaming businesses will collapse leaving just a couple of companies standing.  And that scenario would almost certainly be worse for creators than the current one.

The Steve Jobs Revenue Share Legacy

As I revealed in my book ‘Awakening’, when Steve Jobs struck the original iTunes Music Store deal he walked away a happy man despite having given the major labels the big revenue percentages they wanted.  Why?  Because it meant that it was really hard for anyone without ulterior business aims like Apple had, to make money from selling tracks as a standalone business.  The revenue shares negotiated back then set the reference point for all digital deals since.  The fact that streaming services pay out more than 70% of revenues to rights holders can be traced back to that deal.

The Great Role Reversal And The De Facto Label Monopoly

In the digital era the record labels undisputedly hold the whip hand, and some.  In the analogue era the roles were reversed.  Retailers were the dominant partners and they knew it.  Record labels actually paid retailers for placement to promote new releases.  Compare and contrast that with labels contractually compelling services to provide placement.  Both models are wrong and both engender corrosive behaviour.  Because the major labels account for the majority of music sales it is nigh on impossible for a non-niche music service to operate without all three on board.  This gives each label the effective power of veto.  So even though no major label is a monopoly in its own right each has an effective monopoly power in licensing.  These factors give labels them the strength and confidence to demand terms that would not take place in an openly competitive market.  This, for example, is very different to how digital deals are done in the much more fragmented TV rights landscape.

Loading The Risk Onto Music Services

Why all this matters for the sustainability of streaming services is because of how it manifests in commercial terms.  Recent contract leaks have revealed to everyone the details of what insiders long knew, that labels and publishers front-load deals.  Services both have to pay large amounts up front and agree to guaranteed payments to rights owners regardless of how well the service performs.  (Some labels proudly state they don’t charge advances but instead charge a ‘set up fee’ for every track in their catalogue. Call it what you like, making a music service pay money up front is an advance payment.)  Even without considering the entirely intentional complexity of details such as minimas, floors and ceilings, the underlying principle is simple: a record label secures a fixed level of revenue regardless, while a music service assumes a fixed level of cost regardless.

Labels call this covering their risk and argue that it ensures that the services that get licensed are committed to being a success.  Which is a sound and reasonable position in principle, except that in practice it often results in the exact opposite by transferring all of the risk to the music service.  Saddling the service with so much up front debt increases the chance it will fail by ensuring large portions (sometimes the majority) of available working capital is spent on rights, not on building great product or marketing to consumers.

Skewing The Market To Big Tech Companies

None of this matters too much if you are a successful service or a big tech company (both of which have lots of working capital).  Both Google and Apple are rumoured to have paid advances in the region of $1 billion.  While the payments are much smaller for most music services, Apple, with its $183 billion in revenues and $194 billion in cash reserves can afford $1 billion a lot more easily than a pre-revenue start up with $1 million in investment can afford $250,000.  Similarly a pre-revenue, pre-product start up is more likely to launch late and miss its targets but will still be on the hook for the minimum revenue guarantees (MRG).

It is abundantly clear that this model skews the market towards big players and to tech companies that simply want to use music as a tool for helping sell their core products.   Record labels complain that they don’t get enough value out of big companies like Google and Samsung, but unless they make the market more accessible to companies that are only in the business of selling music they can have no room for complaint.  The situation is a direct consequence of major label and major publisher licensing strategy.

Short Termism And From Evil To Exceptional

Matters are compounded by an increasingly short term outlook from label licensing divisions, with the focus on internal quarterly revenue targets, or if you are lucky, annual targets.  The fact that much of label and publisher digital revenue comprises guarantees and advance payments means that their view of the digital market is different from how the market is performing.  If our small start up that pays $250,000 in rights payments doesn’t even get its product to market, the rights holders still see that digital revenue even though the marketplace does not.  (One failed music service that didn’t even launch went into bankruptcy owing two major labels $30 million).

This revenue comfort blanket insulates labels and publishers from much of the marketplace pain.  So if/when things go wrong, they feel it later, delaying their response.  There is also a cynicism in much deal making, with rigid templates applied to deals and a willingness to compromise principles if the price is right. The latter point was illustrated by the leaked negotiations between UMG and industry bête noir Kim Dotcom in which former digital head Rob Wells referred to being able to ‘downgrade’ Dotcom from ‘evil to bad’ and then from ‘bad to good and from good to exceptional partner’.  The message is clear, if there is enough money on the table, anyone can be a business partner whatever the implications might be for the rest of the market.

Wafer Thin Margins, Deep Pockets And The Innovation Drain

Current licensing strategy biases the market towards those with deep pockets and fatally compromises profitability.  Once all costs are factored in, a music subscription can theoretically have an operating margin of between 3% and 5%. Though only if it doesn’t invest sufficiently on marketing, customer retention and product innovation. But of course the streaming market is in early growth stage so every service has to spend heavily which means that profitability becomes a hostage to fortune. No wonder Daniel Ek is clear that Spotify is a growth business rather than on a profit crusade.

The market dynamics also create an innovation talent drain.  If you were a would-be start up founder the huge up front costs, non-existent margins, and complex time consuming licensing do not exactly make building a music app a welcome experience.  Building a games app however is an entirely different proposition: you own 100% of the rights, you don’t pay a penny to 3rd party rights holders and consumers actually pay for your product.  Music is already a problematic enough sector as it is without burdening it with a punitive licensing framework.

These are the structural challenges that could yet bring down the entire edifice of the streaming music economy.  The irony is that if Spotify has a successful IPO (sans profit of course) it will trigger a wave of copycat services and investment that will perpetuate the status quo a little further.  But it will only be a temporary delay.  Sometime or another the hard questions must be answered.

What Apple Music’s 11 Million Trialists Really Mean

Apple’s announcement that it hit 11 million trialists to Apple Music just 5 weeks after launching has divided opinion as to whether the milestone constitutes success or failure.  It is probably a bit of both:

  • The glass half full perspective: No other streaming service has converted so many consumers so quickly.  Although it is not a like-for-like comparison, 5 weeks after launching its premium tier Spotify had less than 100,000 paying subscribers.  Apple’s early follower product strategy entails waiting for a sector to be ready for primetime before entering it.  Such scale so quickly suggests Apple has timed its entry well here.
  • The glass half empty perspective: It is a very different market now than in 2008 when Spotify launched.  Streaming music listening on mobiles is widespread and awareness of subscriptions is high.  Also Apple hasn’t actually got anyone to part with cash yet.  11 million trialists represents about 10% of the Apple customers that bought music from iTunes in 2014, so the majority of those consumers have not adopted yet.  Also Apple has a well-earned reputation for getting the majority of its iOS users to adopt iOS updates quickly.  So the vast majority of iOS 8.4 users have not yet adopted.

Just How Many Subscribers Will 11 Million Trialists Turn Into?

The biggest issue with the milestone is that we are talking about trialists not subscribers.  So these numbers are an early indicator of likely performance rather than performance metrics in their own right.  What Apple understands keenly – for all its anti-freemium vitriol – is that prolonged free, and near free, access is key to driving subscription.  Just ask Spotify.  So Apple’s conversion rates should be reasonably high, especially considering subscribers are ‘auto opted in’ to payment (and it’s not exactly easy to opt out).  That billing mechanism should ensure more than half of the 11 million make at least one payment.  About half of those will cancel when they realise they are paying, which will leave in the region of 3 million long term paying subscribers from that base, perhaps a few more.

The clever bit for Apple is that the total number of reportable subscribers, the ‘ever subscriber’ number, will be around 6 million because of those extra 3 million that accidentally paid for 1 month.  And because Apple Music should be growing at pace over the next 12 months, we’ll never even see the impact of that 3 million being wiped off the subscriber count, with them quickly replace by new additions.

Where Will Apple Be By Year End?

The 100 million paying subscribers by end of 2015 number is not going to happen though.  And Apple has been smart enough to ensure this number has plausible deniability.  Here’s where Apple looks like its tracking to for the end of 2015:

  • 5.2 million new trialists a month: Following the launch surge the number of new trialists will fall to c.170,000 a day from the current 315,000 a day. Over the course of the year this number will ebb and flow based on marketing activity etc.  This will include 1.2 million monthly family trials
  • 41 million trials by year end
  • 8.7 million active paying subscribers: This assumes a (modest) 5% monthly churn rate and the 25% long term subscriber conversion rate

That will be a highly creditable number in the streaming market though far south of the 100 million.  However this is where Apple can get clever with its numbers.  If it uses the ‘ever subscribed number’ that 8.7 million transforms into 17.5 million.  And if it factors in an average of 4 million users per family account it will be able to report that 69 million people have tried the trial by year end.  Which isn’t so far off the 100 million.

Apple’s got off to a great start in terms of market performance but probably not much more than a solid start by its own high standards.  Even at these rates though Apple is on track to be the number 2 streaming subscriptions provider after little more than 6 months in the game.

User Centric Licensing: Making Streaming Work For Everyone

Artist income is one of the most pronounced growing pains of the streaming era.  While there are many contributory factors, such as transparency and non-distributable label payments, the most significant element by far is how much artists get paid.  There are many moving parts to the equation, not least of which is how much labels themselves choose to pay artists, but even if labels doubled their payments to artists (which would be a good starting point for artists on 15% deals) the underlying dynamic would remain unchanged.  Namely that consumers are switching from buying music (which generates large upfront payments) to accessing it (which generates smaller payments spread over a longer period that as things stand look like they could still add up to smaller amounts even in the longer run).  If you’re a big super star artist or a major label this doesn’t affect you much as you get such a large chunk of the headline revenue.  But a new approach is needed for the rest.  Enter stage left the case for user centric licensing.

Under the current licensing model artists get paid on an ‘airplay’ basis i.e. what share of the total plays across the entire service the artist accounts for.  This model can skew the revenue balance to the superstars who will get played by a very large share of the user base of a service.  Under a user centric model an artist would get paid based on the share of an individual’s listening.  So if a user spends half their time listening to an underground techno producer, half of the royalties go to that producer.  In the existing model that producer would only get a tiny fraction of the royalties generated by that user.

user centric licensing

Let’s take a look at how this could work (see figure).  If a subscriber listens to Artist B 55% of the time but that artist only accounts for 0.5% of total listening, only 0.5% of the available royalties for that subscriber make it back to the artist.  Whereas Artist A who the user didn’t listen to at all gets 10% of the royalty income.  But in a user centric licensing model the artist would get 55%.  The revenue changes from a paltry $0.004 to a more meaningful $0.49 (assuming a 15% royalty share from the label).  And Artist A gets a fairer zero income for zero listening from that user.

Make no mistake, this model will be very difficult to license and the vested interests would likely resist it.  But until we get to scale with subscriptions, we need to explore all ways of ensuring revenues are distributed on as equitable a basis as possible.  This approach won’t fix all the artist-income ills of streaming but it will help smooth the transition.

I’m not going to pretend to take credit for this concept, it’s been quietly gaining momentum for some time now and the Trichordist has been building the case too.  But now is the time to really start giving this approach some serious consideration.  And if the incumbent streaming services are unable to implement user centric licensing because they are too close to the superpowers, then this is an opportunity for a new streaming service to seize the initiative and start to make some meaningful change.

I’m attaching the excel of this model so please go and stress test it yourself. Let me know your thoughts below.

MIDiA Research – User Centric Licensing Model

Making Free Work (Hint Cannibalize Radio Not Sales)

2015 started with freemium fighting for its life. 7 months in and it’s still alive and well but the free debate rages on. It is clear that some form of free experience drives paid subscription uptake but it is also clear that too much free reduces the conversion opportunity. A one month trial is probably too little but a year of free is too much. 3 months is emerging as the free, or close to free, sweet-spot as evidenced by Apple’s 3 month free trial and Spotify’s 3 months for $1 a month. In fact Spotify’s cheap trial strategy underscores the constrained ability of unlimited free to convert to paid. Free is crucial to ensure the acquisition funnel is filled but a new approach is needed, one that is more sophisticated than simply stating it is all free or no free.


One of the biggest concerns about free streaming is that it cannibalises sales. Just for the record, it undeniably does. At least on-demand free does. Free has always been part of the music industry, mainly in the form of radio. But the crucial difference with radio is that listeners do not choose what they are listening to. Free streaming needs to start behaving much more like radio, to follow the Pandora model. Crucially it needs to compete head on with traditional radio. Radio is a $46 billion industry globally yet less than 10% of that flows back to labels and publishers, and then on to artists and songwriters (see figure). By contrast the majority of music sales flow back to rights holders and creators. So the music industry needs to optimise streaming to cannibalise radio more than it does sales. To make the majority of free streaming only partially on demand.

The number one streaming metric that the music industry should be paying attention to is the share of total radio listening time that Pandora accounts for. The more that that increases, the more direct revenue flows into the industry.

free decision tree

But at the same on-demand time free streaming’s role in converting subscribers must be protected, albeit within very strictly defined parameters. Subscribers have two key user journey entry points: 1) a trial 2) free. Streaming services need to make better use of their analytics (which are increasingly sophisticated) to identify which free users to invest time and effort into trying to convert and which to side line. Neither Spotify or Deezer is in the business of free music, they are in the business of subscriptions and simply use free as a marketing tool. So they have no reason to cling doggedly to free users that show no sign of converting. Instead after a sufficient period of free music has been offered users should be pushed to subscriptions or onto a radio tier (see figure). There is no business benefit to the streaming services nor rights holders to have perpetual on demand free users.

The assumption that free music is some sort of internet right is symptomatic of the internet’s growing pains. In terms of market development we’re probably at the adolescence stage of the internet, the stage at which carefree childhood starts to be replaced by responsibility and consequences. We’re seeing this happen right across the internet economy, from privacy, data, free speech, jurisdiction etc. Because music has been free online for so long consumers have learned to accept it as fact. That assumption will not be changed any time soon, and try to force the issue too quickly and illegal services will prosper.

Of course YouTube is, and always has been the elephant in the room, buoyed by the schizophrenic attitude of record labels who simultaneously question its impact on the market while continuing to use it as their number 1 digital promotional channel. While the tide may finally be beginning to turn, don’t expect YouTube to go anywhere any time soon. But should the screws tighten do expect YouTube to stop playing ball. As they have made clear in various rights holder conversations, an onside YouTube, warts n’ all, is far more appealing prospect than a rogue YouTube. But implicit threat or otherwise YouTube must be compelled to play by the same rules as everyone else. As I’ve said before, YouTube needs to look more like Pandora.

Competing against radio needs to become the modus operandi of streaming. Only when free music on the internet evolves to more closely resemble radio will the industry be able to fix the apparent paradox of increased consumption translating into reduced revenue.