Sixteen years after the music industry's peak, revenues have returned to growth. But the core problems of streaming service profitability and minuscule artist royalties persist. There is cause for optimism, but transformation is needed. Enter, Spotify and Apple music groups?

Brand new Beretta, can’t wait to let it go /
Walk into my label like ‘Where the check tho?’
– Drake, “Star67”


After 16 consecutive years of decline, the music business finally started to grow again in 2016. Year-over-year, recorded music sales were up 11% on an inflation-adjusted basis, worth $641MM. This feat was almost exclusively thanks to subscription music services, which saw revenue growth of 99% and, at $1.5B, represented nearly 250% of net gains. During the following year, revenue grew an astonishing 21% (or $1.4B), with subscription revenue up 59%. Not only are these trends expected to continue in 2018, one of the year’s largest and most anticipated IPOs was in the business of selling recorded music.

Sales of recorded music may still be down 63% since its peak in 1999, but the return of growth is great news no matter who you are or how you interact with the music business. But it’s especially good for the major labels. Not only do they collect the majority of streaming royalties, this revenue is higher margin than ever – there’s no need for traditional costs and risks such as album manufacturing and fulfillment. Even better, an unprecedented share of revenue is coming from music that was recorded (and paid for) decades ago.

redef_music2_1-1Yet, focusing just on renewed topline growth overlooks many of the sustained problems affecting the music ecosystem: Too many artists remain unhappy and poorly paid; the major streaming services remain widely unprofitable, with several having recently folded and many others circling the drain (and who knows what’ll happen as the major labels continue divesting their stakes in these companies); a substantial portion of music consumption remains un/under-monetized; and rights conflicts continue to proliferate.

The best way to address these issues is to fundamentally rethink how the music industry should be organized – or asking, “If you were to recreate the business from scratch, how would it work and how would it differ?” To do so, it helps to consider the industry in waves. If Wave 1 was the shift from live music to predominantly pre-recorded radio play, Wave 2 would be the shift from radio to recorded for sale, with Wave 3 representing the shift from physical to digital purchases. Wave 4 is our current transition from digital purchase to streaming. The first wave fundamentally changed the way the industry worked (e.g. more formalized managerial layer, national stars, touring etc.), while the second created the label system we know today. During the third wave, the industry changed in response to declining revenues (less A&R, fewer artists, 360 deals) but the players and how they were structured remained largely the same – the primary response was consolidation in pursuit of lower costs, greater bargaining power and more efficient A&R. By and large, the constructs that supported waves two and three persist as we embrace the fourth. The labels haven’t really changed – they still focus on the same core activities, provide largely the same artist revenue shares, same model of advances and recoupments and so on. The incongruency of this stagnation is significant. While the third wave is typically seen as the most disruptive, the fourth brought about two revolutionary changes to the economics of music. For the first time, the monetization of music was directly associated with its consumption and consumers could listen to as many artists or tracks as they liked at no incremental cost. Despite this, the most significant disruption at the labels has been who sends them the checks.

There’s an irony here: Though music was one of the first mediums to go digital, the business is still re-applying the structures of yesteryear onto the needs and complications of today. Companies that have tried to innovate, such as SoundCloud, have been stymied almost out of existence (with the survivors, such as Pandora, now years behind their original trajectory). If the music industry wants to solve its troubles, things need to change – and that means understanding what’s wrong.


Artists, Composers and Producers
The biggest challenge facing creatives in the 21st century has been commoditization. While we’ve historically believed creative aptitude was exceedingly rare, this wasn’t quite the case. There’s an abundance of talented vocalists, musicians and writers – it’s just that they were rarely discovered, and fewer still were produced, then distributed, then marketed. We saw this commoditization with the shift to flat pricing for digital downloads – all songs were suddenly worth 99 cents.

Many in the music business regretted this move and, accordingly, saw streaming as an opportunity to re-introduce discrimination. However, growth in interactive radio or curated music services has been modest in comparison to on-demand services like Spotify and Apple – especially when adjusted for usage (Spotify averages roughly 3x the listen time per user of Pandora). Unlike Netflix (~5,000 titles) or Pandora, the value proposition of these services is based on having all (or roughly all) the music a user might want. And this has a defining effect on how rights holders are paid, as individual services can’t acquire licenses to individual catalogs or titles at scale – it would be contractually impractical and analytically infeasible. After all, what’s the value of one individual stream in a month of 1,500+ streams (Spotify’s per-user monthly average in 2016)? And if that track hadn’t been available to stream, how much less valuable would the service have been to the consumer? Would they have unsubscribed? And how might this value change over time (i.e. immediately after release v. a year later)? As a result, Spotify establishes the “value” of every track in a consistent, straightforward fashion: what share of total streams the track represented. This means that if Drake’s music represents 10% of total plays in July, his music will receive 10% of July’s total revenue-based royalties. This codifies the idea of a commodity: Every unit is worth the same price and revenue is mostly dependent upon volume.

Though this is a simple way to value creative product, it’s also the most practical at scale. It also rewards lasting music over flash-in-the-pan hits with limited staying power. In the era of purchased music, any two transactions generated roughly the same revenue for the artists responsible, even if the lifetime value of those purchases varied dramatically (e.g. an album you bought and regretted v. the album that defined your high school years). Millions bought Psy’s viral hit “Gangnam Style” for $0.99 in May 2012, for example, but how many listened to it a month or year or five years later? It didn’t matter. In the Spotify era, each track generates revenue the more it’s played. If you continue to play Sgt. Pepper’s Lonely Hearts Club Band, the Beatles keep making money. Don’t play “Gangnam Style”? Psy doesn’t make a dollar. It’s hard to say this dynamic is unfair to the artist, or unhealthy for the industry.

However, this model does have its problems. First, per-stream rates will, by mathematical necessity, seem paltry. Divide a $10 monthly fee by 1,500 monthly streams and you get a “value” of $0.007 per stream. This is especially true when compared to historical measures, such as unit sales (which effectively represented paying for all future listens upfront) or radio plays (where each play reached thousands of listeners). Making matters worse, the royalties that individual artists ultimately see are first reduced by the label’s share, and then split across each of the artists responsible for a track (i.e. the composers, performers). Second, these rates will continue to decline as per user engagement increases (if a user moves from 1,500 tracks per month to 2,000, ARPU remains flat but their effective per-stream rates fall 33%). Third, per-stream rates rapidly oscillate. The same number of streams may deliver more or less revenue from one month to another (as “value” represents the share of total streams). Fourth is hyper-fragmentation. If you were to spend $120 in 2002 (equivalent to a year of Spotify at $9.99 per month), this could buy only five to six albums (and thus compensate five to six artists). In the on-demand streaming world (enhanced by algorithmic playlists like Spotify’s Discover Weekly), this spend is spread extremely wide. More artists are paid, but they’re paid far less on average – it’s one thing to be one of five albums bought in a year, another to sustain 20% of a listener’s on-demand streams over that entire year.

Another key issue is the timing of streaming income. Major artists are used to collecting the majority of the revenue from a new album shortly after its release, as most sales are frontloaded. However, streaming royalties are paid only when and to the extent that listening occurs. While this means that superfans could generate more income from an artist (v. buying a single copy of an album or track), it also results in trading a large upfront payment for a higher potential (but not guaranteed) income stream. As a result, the financial success of an album may not be clear for several quarters or even years and the size of the first check will be far more modest. While this rewards music with legs, it will test the patience of those who feel their bank statements lag their stardom.

All of this creates a particular problem for artists – but not for labels. As the labels collect a fixed percentage of a streaming service’s total revenue (typically 70-75%, before paying out to artists/songwriters/publishers), they’re not affected by unit rates or month-to-month oscillations. Their aggregated view means they’re also less affected by fragmentation among their artists – total revenue remains the same  Given their exposure to countless artists and extensive back catalogs, the labels also don’t need individual artist revenues to be front loaded. All they need is aggregate stability.

The primary problem, however, is how the major labels monopolize royalty payments. Spotify and Apple Music take roughly 30% of total revenues (which goes to operating costs, as well as customer sales tax and platform fees), with the remaining 70% paid out in royalties. Out of this remainder, the major labels keep roughly 70%, with 15% going to performers and 15% to composers. And remember, a hot song often boasts a handful of writers and several performers, each of whom will share in the net royalty (Spotify’s most streamed track in 2017, Ed Sheeran’s “Shape of You,” counts six writers; Kanye West’s 2015 hit “All Day” had four performers and 19 credited writers).

REDEF_MusicEconomics_1.13Almost every media business today is contending with declining or flat inflation-adjusted revenues and compressing margins. Yet major music labels continue to take the same cuts they have for years – even though they do less than ever to generate a dollar in revenue or to get music into audience’s ears (in a streaming environment, their 70% cut isn’t needed to cover manufacturing and shipping). To put this in context, consider artist complaints about per-stream rates (even if this criticism is overblown). This can’t be fixed by doubling the cost of a monthly subscription (which would likely reduce total subscription revenue) or increasing licensing fees from 70% to 100% (obviously impractical). You can’t solve the rate problem without addressing label’s 70% cut of total royalties.

A common rejoinder to this argument is that growth in subscriptions will solve the problem – if everyone had Spotify or Apple Music, per-stream rates would remain low, but gross payments would increase substantially. There are three limits to this argument. First, prices would likely need to drop in order to drive additional penetration. In fact, they already are as the major services embrace student pricing and family plans (which cost 50% more but allow four to six unique accounts): Over the past three years, premium user ARPU has fallen from $7.06 per month to $5.25. To this end, family plans exert significant downward pressure on per-stream rates, as the number of streams grows substantially more than revenue. For related reasons, the industry is also unlikely to return to the days where the average American over 13 spent $80-105 a year (1992-2002). Even if every single American household subscribed to Spotify or Apple Music, per capita spend would be around $65-70. This is still more than twice today’s average of $31, but such penetration is unlikely (in 2017, only 80% of American mobiles were smartphones). Put another way, much of the remaining growth in on-demand streaming will come from adding additional users to existing subscriptions. While this increases total revenue per subscription (from $120 to $180), it drops ARPU to at most $90 and its lowest, $20.

Second, growth in on-demand music subscriptions is likely to cannibalize the terrestrial and satellite radio businesses. In 2017, SiriusXM (which has the highest content costs per listener hour in the music industry) paid out $1.2B in US royalties, roughly 33% of that of the major streaming services. US terrestrial broadcast revenue generates another $3B+ in annual royalties. These formats are rarely considered when discussing the health of the music industry, even though one reflects direct consumer spend. But they provide significant income for the creative community (though notably, terrestrial radio royalties compensate only composers, not performers). As on-demand streaming proliferates and cannibalizes more terrestrial/satellite radio listening (still more than half of total audio time in the United States), streaming royalties will continue to grow – but much of this will come at the expense of radio royalties.

Most important, relying on total revenue versus unit revenue eschews the question of fairness: A huge portion (if not the majority) of value is created by artists and the streaming services, but most of the proceeds are captured by the labels. As a result, we’re seeing increased efforts by major artists to escape, reshape or avoid major label deals. In 2015, Drake released the pointedly titled mixtape If You’re Reading This It’s Too Late in order to fulfill obligations under his four-record deal in advance of his next album, Views (which sold more than 4x the copies of IYRTITL but under a new deal under UMG). The same year, Frank Ocean was even cheekier: releasing the “visual album” Endless through Def Jam one day before releasing, now as an unsigned artist, Blonde. Others, such as Macklemore (who remains unsigned following his hit album The Heist) and Taylor Swift (represented by Big Machine, an independent co-owned by her father) continue to rebuff major label offers – even though these artists hold most of the bargaining power. Instead, Macklemore and Swift (via Big Machine) hire WMG and UMG to manage distribution. And in 2016, Chance the Rapper self-released a streaming-only album, Coloring Book, that proved that neither label support nor physical media were necessary for success. The album hit #8 on the Billboard 200 albums chart and was nominated for Album of the Year and Best Rap Album at the 2017 Grammys, winning the latter.


Streaming Services
The music streaming services face a core problem: They don’t make money. This isn’t a question of scale. As they’ve grown their revenues and subscriber bases, losses have worsened. This strategy isn’t unique in tech, but it is unique for media distributors. “In every other media business I cover,” Piper Jaffray analyst James March told Quartz in 2014, “if you sell more tickets, or get higher ratings, the operating leverage is massive, you don’t pay more [in licensing fees or royalties] for the TV show or movie. But in streaming music the more popular they get, the more royalties they pay.” Netflix, for example, essentially pays fixed fees for its content. Regardless of how much revenue Netflix generates (or doesn’t generate), this licensing costs stay flat – creating the opportunity for outsized profits, as well as losses. Spotify and Apple pay out at least 70 out of every 100 cents in revenue generated. And based on technology and marketing costs thus far, their 30% cut doesn’t seem to be enough. Over the past three years, Spotify has averaged an 11% operating loss before accounting for debt (which takes net profit to -30% of revenue in 2017, up from -18% in 2016).

redef_music2_1-3These tribulations have helped drive the music services toward non-music content – namely podcasts and video. While these boast fixed licensing costs (and are thus good for the services), they de-emphasize musical artists and stand to cannibalize their revenue. And while the likes of Spotify and Apple Music have restored growth to the music business, is it truly healthy if it needs other mediums to survive? Similarly, each of Spotify’s primary competitors is willing (if not dedicated) to taking a loss in the segment for strategic reasons. Amazon Music Unlimited, for example, costs only $7.99 if you’re a Prime customer, with Amazon taking a $2 discount on an a business with only $2.0-3.0 gross margins. In November 2017, Apple Music’s Jimmy Iovine warned against optimism in a Billboard interview: “The streaming services have a bad situation, there’s no margins, they’re not making any money. Amazon sells Prime; Apple sells telephones and iPads; Spotify, they’re going to have to figure out a way to get that audience to buy something else.” To point, the other standalone services have largely disappeared. Tidal’s backers, which include the crème-de-la-crème of musicians, have been looking for exit for close to three years; Pandora seems to have receded to ad-supported interactive radio; Deezer has been essentially stagnant since cancelling its IPO in 2015, and SoundCloud has refocused on B2B revenues and community/amateur music.

Spotify, despite its leadership position, faces two key challenges as an independent. Apple Music continues to pick up momentum and is expected to surpass Spotify in paid US subscribers by summer 2018. What’s more, REDEF expects the company to soon bundle Apple Music into other Apple subscriptions, including iCloud, Video, Apple Care and even device purchases. In addition, Spotify is still partly owned by the three major labels, but two divested a substantial portion of their stakes two months after its IPO (Warner sold 75%, Sony 50%, with Universal yet to announce any sales). One has to wonder what will happen to the company as it loses the support of these VIP benefactors. It’s hard to see this as a positive change, as it moves the relationship from being one of compensation by both equity appreciation (a non-cash cost to the services) and license fees (a cash cost) to only the latter. And if/when Spotify does start to generate a profit, it’s reasonable to imagine the labels will slowly ratchet up their rates to collect much of the take. Apple – the most valuable and cash-rich company in the world and one using streaming music to fuel that business, rather than being a business in and of itself – may also find itself a victim of its own success. Why wouldn’t the labels push up their rates as penetration-related revenue growth slows?


Major Labels
Despite rising revenues and margins, the major labels face serious challenges. On the one hand, they must contend with escalating disintermediation by the streaming services. And on the other, the fact that these same services are making much of a label’s core functions obsolete. Historically, labels were critical to being able to reach an audience – someone had to manufacture and press a record, convince retailers to sell it, manage shipping and fulfillment so consumers could buy it, and ensure that audiences were aware the album existed in the first place. Today, artists can skip most of these steps through either direct-to-consumer distribution or streaming-only releases. And if an artist does still decide to use a major label, an increasing share of total listens will still occur on streaming services – in which case the labels are taking a massive cut despite providing significantly fewer (and lower cost) services. What’s more, labels’ ability to drive (or force) consumption in the on-demand streaming era is a fraction of what it was in the age of radio. Promotion and radio play still matter, but they’re less essential (and impactful) in an era of algorithmic recommendations (e.g. Spotify’s Discover Weekly), on-demand selections, playlists and artist-to-fan social sharing. Even A&R has been diminished in an age of YouTube stars and the broader democratization of taste. This overall dynamic isn’t sustainable and creates either an opportunity for disruption or suggests a far more modest margin future.

However, the labels do have a key advantage in the digital era. As cross-platform suppliers, they can mix data from every streaming service (e.g. Apple Music, Spotify, YouTube, Amazon) to provide aggregated insights and revenue information to artists. The product engineering culture and technical skills required to support this type of service offering are different than those resident at the labels today, and they’re not easy to get right. And to point, most artists will say these tools remain woefully inadequate as streaming approaches two thirds of total consumer spend on music. The streaming services frequently complain that even foundational output data – who wrote a song, who published it, who owns which rights – is unreliable and often conflicts between copies. To succeed, the labels will need to not just aggregate and summarize data across services, but generate actionable insights to drive usage.

To be fair, the labels have never needed this data infrastructure before. However, the shift to digital means the labels are essentially in the rights allocation business; if they want to thrive, they need to excel at this capability. Still, even a fully-baked audience analytics system may not be enough to earn significant cuts of royalties going forward. If we move to a “winner takes most” environment, with Spotify and Apple Music capturing 80% of on-demand streams, the need for a cross-platform data service will be diminished. Not only are each platform’s artist tools likely to be sufficient, most analysts assume they will be more detailed and sophisticated than those published by the labels. As such, the future of the major labels may depend on something they can no longer control – but spent years indifferent to – the number of viable, large-scale direct-to-consumer streaming music services.



Envelopes coming in the mail, let her open em’
Hopin’ for a check again, ain’t no tellin’
– Drake, “No Tellin’”


Artists, Composers and Producers
Revenue growth is obviously good for artists, but neither this growth nor radical changes in streaming economics (e.g. doubling prices to $19.99, increasing royalty rates from 70% to 90%) will solve the aforementioned problems. To do so, a total change is necessary – one in which the steps between the artist and fans are minimized (though this doesn’t necessarily mean D2C), label agreements and services are rewritten and copyrights are retained by artists.

redef_music2_1-4Whether it’s 1950, 2000 or 2050, artists need infrastructure, systems and support to develop, release and monetize their music. However, this need no longer require giving up 50%+ of revenue in perpetuity, handing over copyrights, or audience/data disintermediation. Accordingly, we’re likely to see a growing shift to hiring label services on fixed fees for specific tasks. U2, which had 2017’s top-grossing tour, is a large operation, to be sure, but its also a reliable and routine one. Why can’t the band pay a salaried team to manage its tours, contracts, studio sessions and retail distribution (the last of which Macklemore hired Warner Music Group’s Alternative Distribution Alliance to manage on The Heist)? And if indie artists can have a multiplatinum record without label support, what must the Katy Perrys and Beyoncés of the world think about collecting less than half the proceeds of their own music? Label responsibilities, such as retail negotiations and managing manufacturing and fulfillment contracts, are becoming anachronisms in the age of digital distribution and flat prices. Conversely, the importance of brand and IP management is more important than ever in the era of 360 deals, celebrity venture capitalists and direct-to-fan engagement. Artists need to own their image, as well as when and how they release their music. This necessitates either significant changes in the label model or a new model altogether.

There are some leading companies addressing these problems. Kobalt Music Group, for example, provides publishing and label services (without taking copyrights) and has offered artists direct access to richly featured data dashboards for years. UnitedMasters is another interesting example. The company leans heavily into CRM and audience targeting tools that cross not just multiple music platforms, but also non-music platforms like Facebook in order to drive ticket sales and merchandise. UnitedMasters positions itself as more of a growth-support company than a label. Warner Music Group’s Alternative Distribution Alliance (founded in 1993), meanwhile, provides independent artists and labels simpler and low-cost access to physical and digital distribution resources, in addition to services such as music licensing, merchandise and sponsorships. Increased royalty rates, the retention of copyright upside and better analytics are great for artists. However, the services afforded by these companies are limited by the fact they lack direct relationship with consumers, own little of the data used to support their business, and cannot directly affect consumption – but there are those who boast all three.


Streaming Services
The major services have three advantages. First, is their growing scale (defined by number of users and share of industry revenues) and the fact that the market seems to be organizing itself into a duopoly (Spotify and Apple Music). In time, this should produce additional negotiating leverage with the labels. That said, the streamers’ bargaining power is limited by the fact their value proposition depends on having every major label (each of which drives 20-30% of total consumption). Pricing is another opportunity. As penetration and engagement grow, the services have the opportunity to continually raise prices (as Netflix has). While flat gross-margin rates would still net the services only 30% of the proceeds, an additional dollar in ARPU would have offset Spotify’s annual loss in 2017. What’s more, this price improvement need not come from increases in standard monthly fees (though this is likely). ARPU can also be increased by reducing the implied discounts for family plans (particularly relevant as users-per-account grows) or pulling back on promotional pricing (e.g. student discounts). These efforts may be stymied by competitors such as Apple and Amazon, both of which uses music as a loss leader rather than a business. Still, neither reducing rates nor increasing pricing may be necessary. Topline revenues will continue to grow and although gross margins will remain flat, gross profit dollars will still scale. The largest services are likely to achieve operating leverage as marketing and R&D costs, in theory, stabilize.

However, the services’ most significant potential comes from two other advantages: direct customer relationships and behavioral data. When these are applied at scale, the services have unprecedented ability to help artists grow, to understand what works and what doesn’t, quantify (and identify) audience passion, drive awareness, engagement and revenue, and so on. Apple Music and Spotify are rapidly building out these capabilities, but their ability to leverage (and monetize) them is limited by disintermediation by the only member of the value chain steeped in cash: the labels. This won’t last.

Streaming services have an opportunity to cut out labels by forming direct-to-artist deals or establishing their own pseudo-label services. Not only has this long been predicted, it’s been incubated for years. Since 2015, the major services have cultivated exclusive windows and radio shows with major stars, including Beyoncé, Kanye West and Drake. While this construct still went through the label system, it generates clear business cases for further disintermediation.

The most transformative changes will go beyond replicating the label model or snapping up top artists as their deals expire. Spotify and Apple are likely to use their customer data to identify and test up-and-coming artists and predict future performance for major artists, then offer advances in exchange for reduced royalty rates on a defined number of streams (to recoup) and then a permanent royalty rate that, while lower than the gross rate paid today, would provide the artists with significantly more than they receive after a label’s cut. For example, Spotify might offer Drake (who would collect ~$0.0015 per stream today after label fees) $10MM in exchange for a permanent royalty rate of $0.0035 per stream (roughly half of Spotify’s per-stream royalty rate), with 0.002 paid across the first billion streams. Assuming five billion incremental streams (Drake is approaching 15B on Spotify alone), Spotify would save $14MM and Drake would make another $16MM (though his costs would go up). And that’s before considering availability on other services, such as Apple Music, which would generate Drake more incremental revenue, as well as generate cash for Spotify (more on this in a bit). The services would also give artists full access to their data dashboards, analytics and other services (e.g. ticket sales). And most important, all copyrights would stay with the talent responsible.

The streaming services would be able to do similar (but smaller) deals with artists with early signs of breaking out (Apple’s $400MM acquisition of Shazam is undoubtedly part of this plan). Not only does this solve economic issues for both the streaming services and artists, it’s scalable. As total streams (i.e. subscriptions and engagement) increase, the services can spend commensurately more on these advances. And as these deals grow, the cashflow impact will subside (i.e. the services will be paying out advances while benefiting from a substantial reduction in royalties). The deals also create incentive alignment between artists and the services. Consider the following research from AWAL, a streaming-only label owned by Kobalt, which shows how inclusion in one of Spotify’s featured playlists affects total plays – both through playlist listens and driving increased awareness.

redef_music2_1-5This type of move is likely to produce significant backlash from the major labels, who would see it as an existential threat. But with streaming services driving all revenue growth and close to half of total revenue (with most of this revenue concentrated among just Spotify and Apple Music), the balance of power will continue to shift from labels to the major streaming players. If the labels try to pull their music off the services in protest, they’ll face tough conversations with their biggest stars when explaining the income shortfall and upset fans. Accordingly, the labels will need to prove their value add and economics in the on-demand, digital streaming era. They’re undoubtedly better at growing artists into stars, but this comes at a significant cost that could be, at least partly, replaced with fixed-fee services rather than waterfall equity. And we’ve yet to see what the streaming services can achieve by building (or poaching) A&R teams and providing them with full access to their analytics engines and enabling them to perform audience tests (e.g. A/B testing artists in Discover Weekly to assess their audience potential before signing any royalty agreements, or testing various mixes for audience preferences).

There is also the question of whether an artist signed to Apple would be available on Spotify, and vice versa. For major artists, the answer is likely to be yes. Not only would the financial cost of forgone streams be significant, these artists require mass market fandom and attention to fuel other income streams (e.g. touring, merch) and maintain their cultural cache. And neither service benefits from balkanization. Not only would a self-centered focus undermine their sales pitch to artists (v. labels who would want to maximize direct revenue for all parties), it would fragment consumption data, erode their value propositions (access to all the music you could want) and hinder adoption of streaming services (why bother if you need to pay two). However, the services are likely to sign up-and-coming artists on exclusive deals. Not only do they provide a low cost, low risk opportunity to accumulate a differentiated offering (while avoiding strategic blowback), the pitch to indie artists (money + discovery for exclusivity) is compelling and allows the services to use their data to compete against the labels and one another (versus just cutting checks).



The future for the major labels is daunting. A&R is still a value-add function, but burgeoning artists no longer need to go through gatekeeper scouts and producers to reach audiences and develop followings. Fans are increasingly responsible for artist discovery. Production continues to decentralize while also getting cheaper. And as more listening hours shift online, traditional label activities such as manufacturing, distribution and promotion become less important. Instead, labels will need to be in the rights distribution and data analytics business. The major labels aren’t great at this today and it’s unclear whether they can build a more compelling data offering than the services. Without a platform, a meaningful direct-to-consumer relationship or ownership of end-user data, opportunities will be limited.

In the short-to-medium term, the labels will still benefit from substantial growth in streaming revenues – most of which will cost the labels nothing to generate. In addition, the transition to streaming-service labels will take years as the industry (and its behind-the-scenes talent) reorganizes, Spotify and Apple ramp up their investment in “label-like” licensing, and existing record deals lapse. Furthermore, artists will still need more than just advances. Tour booking, planning and promotions, radio play, and media coverage (e.g. talk-show appearances, Rolling Stone covers) all matter – and the streaming services are unlikely to provide this support in the early days, or even en masse down the road. In addition, rights policing (e.g. unauthorized YouTube uploads, soundtrack use, covers, etc.) and licensing has become more important in the digital era, and it’s hard to see the streaming services providing much support here (especially given they wouldn’t own copyrights).

Over time, we’re thus likely to see the major labels transition into catalog administrators for talent they’ve long since produced (this segment is already responsible for upwards of 100% of profitability). They’ll also have a sizable but low-margin side business in artist management and artist services (operating similarly to that of the Hollywood talent agencies) and occasionally serve as from-scratch A&R hitmakers (albeit at much lower revenue share rates). More broadly, labels will come to look more like Kobalt than their present-day selves – and someone (perhaps Spotify or Apple) will undoubtedly purchase the smaller, digital-centric services companies like Kobalt.



Over the past 20 years, the music business has faced more challenges than any other media category – and perhaps any other industry (revenue has declined twice that of the traditional taxicab industry). Some of these challenges were of the industry’s own making, while others were bad luck (music is the easiest content to pirate). With recorded music revenues up 35% in two years, there’s cause for optimism – but disruption and change accumulate over time. In 2018, much of the infrastructure and scaffolding supporting the music business seem ill-fitting, inefficient or unfair. The music industry needs to shed much of its historical structures and processes – not just how music is found or heard, but also who finances it and how. This doesn’t mean today’s giants will go away, but their standard deal terms are unlikely to survive.


You can reach Matthew Ball at or @ballmatthew