The worth of music

I've long been fascinated by the music economy, one of the earliest blogs I ever wrote covered this very topic over ten years ago.

Much has changed in the last ten years. At the time of that original post, only a quarter of Spotify's subscriber base was on a paying tier; as of 2026, this has now grown to almost half its subscriber base. Apple Music was on the precipice of launching, and Apple was lobbying the record labels for permission to set the price of streaming down to $5 - $8 per month to convert more subscribers to paying customers. The labels wouldn't budge on the $10 per month price floor. Today, the cost of an ad-free music streaming subscription has remained roughly the same, or become marginally cheaper in real terms in the last 15 years and we don't talk about it anymore. The streaming economy has seemingly set its equilibrium price.

What hasn't changed though is a perception that artists are being slighted. At the time, I argued that record labels were a glut on an industry that had outgrown them. With the internet and music tools available, artists are free to record, market, and distribute their own music, so what is a record label doing skimming 45% of the revenue off the top? Without their fingers in the pie, surely there'd be enough revenue to go around to compensate artists fairly.

Of course, this is an oversimplification as labels do provide significant administrative, financial, and creative support to artists scaling their careers. It's simply that the extent to which they are crucial to artistic success has certainly diminished. The record label fraternity hasn't collapsed by any means, but true enough, the share of music released that is tied to major labels has dropped notably. On Spotify specifically, the three majors plus Merlin (the indie-label licensing body) accounted for 87% of all music streams in 2017, falling to 71% in 2024. On the other side of the coin, the number of independent/DIY artists dwarfs that represented by major labels - independent and DIY distribution made up 96.2% of daily uploads to streaming services in 2024.

Evidently, modern technology has made it easier than ever for independent artists to create and distribute their music, but it has somewhat counterintuitively been disastrous financially for this long tail. Spotify's own 2024 Loud & Clear report is damning on this point: of roughly 12 million artists with music on the platform, only 71,200 (about 0.6%) generated $10,000 or more, and just 12,500 crossed $100,000.

By its nature, the streaming model benefits large artists with already significant audiences. Streaming subscription revenue is aggregated into a pool and then sub-divided on a pro-rata basis on total number of streams. Some streaming platforms have minimum stream thresholds before any revenue is paid, meaning the pro-rata pool of funds gets redirected towards already higher streaming artists. It's a system that benefits the labels representing major artists because...well it was created by them! Streaming was created to save the labels from piracy, and it did. But it's also created the conditions that now hold back independent artists financially.

None of this data is particularly surprising. Of course the vast majority of music is independently distributed, and of course the industry's revenue is vastly skewed towards the top echelon of artists represented by major labels. Becoming a professional musician is a privilege that most never reach. Creating music is a labour of love and most of us gladly do it for free. How problematic is it then that most musicians struggle to get paid? If we would do it for free anyway, does that make it worth...nothing?

These would be much harder questions to answer if there wasn't precedent. The numbers consistently show that on a per capita basis, in real-terms, the average consumer spent more on recorded music at the industry's peak in 1999 vs today. If we take the RIAA's own historical data, US recorded music revenue was $14.6B in 1999, which is ~$26.9B in today's dollars. In 2024, the same RIAA data reported US recorded music revenue was $17.7B. Per capita, we used to spend nearly twice as much on recorded music in 1999 vs 2024.

So, the pie isn't as big as it used to be. Is music worse than it was 25 years ago that's causing us to spend less? That's probably not the variable. Much of it can be attributed to simple supply-demand economics. The sheer abundance of music is the primary variable. With the tools to create, market, and distribute music available at the edges, there is way more music than there's ever been and the supply of music is growing faster than the number of ears available to consume it. But the internet has also created abundance in a more sinister way by removing virtually all the friction to access and consume music. If the singular macro-level characteristic of the technology - specifically, software - revolution is to gradually reduce all marginal costs of production and distribution to zero, then it shouldn't surprise us that it also inevitably drives all marginal revenue to zero.

This is a big problem for the music 'middle class' if you will. Thirty years ago, music was harder to make and harder to distribute, but making a respectable, middle-class income as a musician was much more attainable. In the CD era, selling 20,000 copies of a $20 album a year at roughly $4 take home per album netted $80,000/year, purely from the sale of recorded music. To take home that same $80,000/year today, that artist would need 16 million streams that year. That's an order of magnitude change in the size of audience needed to earn an income on recorded music.

While this blog is about the music industry, the music industry isn't unique in what it has succumbed to. This same phenomenon is seen commonly across industries and professions that have been touched by the internet. Lamenting the 'hollowed out' middle-class has become highly topical in US politics as the US economy has over decades become reliant on high-leverage services and technology industries that, while nominally productive, aggregate wealth at the centre rather than at the edges. Money gravitates to where scarcity is, and in a world where production and distribution is limitless, what's scarce is discoverability, control, and curation. These are things served by aggregators instead of creators. News, magazines, and journalists are subservient to search engines, content creators are subservient to social media, and musicians are subservient to streaming platforms.

So music is a microcosm of a trend that has gripped many industries. Generative AI is yet another accelerator of this unstoppable trajectory. Spotify has recently removed some 75 million 'low-effort' AI-generated tracks from its library and is implementing features to help users better distinguish AI and human-generated music. But it feels a bit like patching a busted pipe with masking tape. Generative AI is yet another tool that makes music easier and faster to make. We will have more music, and our demand will be spread over yet more content, making it harder for anyone to make a middle-class living. The content machine as it is doesn't reward care or intentionality, it is receptive primarily to volume.

It goes without saying that there are mixed feelings about AI in artistic fields. Tools that primarily intend to reduce or eliminate the arduous process of creation defeat art's very purpose. The process of creation is the act of expression. But creating music is also a profession measured in more rational terms - it's an arrangement of sounds designed to be pleasant, generate value for the listener, and revenue for the maker. If those outcomes can be achieved more efficiently and at a larger scale then it becomes a more viable profession. One could argue that we've always had this dichotomy - there is music that is manufactured for revenue, and music that is made for artistic intent. Is it any surprise that one of them makes money, and the other one doesn't?

History tells us that it's not so black and white, and the long tail of 'independent' music has always had its backers. But as music gets easier and faster to manufacture, and consumption gets more and more frictionless, the long tail of 'independent' music is becoming financially drowned out. In a world where technology demands speed, independent music must also compete on that metric, and by its very nature, is not designed to. The only way to restore the middle class of independent music is to find a safe space that is immune to the immense speed and disposability of the digital economy. More and more, that is looking like a return to physical media.

The resurgence of vinyl and cassette is well-documented, and while this is generally seen as a nostalgic yearning for simpler times, it can also be considered an act of economic defiance. In a world where music is becoming increasingly disposable and cheapened, the marginal cost of recording and distributing something on a physical medium is a hurdle that 'fast-fashion' music won't cross. With the revenue on offer on streaming services being virtually non-existent for the majority of artists, independent artists wouldn't be remiss to eschew the digital streaming paradigm entirely and return to physical media. Yes, distributing on physical media comes with real-world costs, which are a higher barrier to entry than simply uploading songs to streaming platforms. But the revenue available is orders of magnitude greater, and affords artists the ability to own their own distribution outside of an ecosystem that is diluted and no longer working for them.

Perhaps an underwhelming conclusion to this blog is that the worth of music is merely a product of supply and demand. In the last 30 years, music has been subjected to forces from every direction that have vastly increased the volume and accessibility of music. Music is worth what it is, and there's only so much to go around. The pie itself has gotten smaller by virtue of music's sheer availability, such that consumers have had the bargaining power to spend roughly half as much money for music per capita than we did 30 years ago. But what is even more problematic is the decimation of the music industry's middle class. The shrunken pie has mostly shrunk the share enjoyed by the industry's long tail of mid-tier, independent musicians. This can mostly be attributed to the way streaming services are designed which benefits music at two extremes - major artists with massive reach, or low-effort 'fast-fashion' music - at the expense of a now 'missing middle'.

A viable life-raft is for independent artists to opt out entirely. The worth of music is something that can be demanded, by rejecting the mediums that devalue it.