Introduction

Do you have gaps in understanding how the music industry works - royalties and revenue streams like mechanicals, performance royalties, and sync licensing?

The best way to get a clear picture of the inner workings is to explore each major technological advance and then look at how intellectual property laws evolved to shape and monetize the music industry.

In Part One we covered music publishers and songwriters collaborating to print music and collect public performance royalties.

In Part Two we covered audio recording technology, mechanical licenses and radio broadcast royalties.

In Part Three we covered film and television, the role of music supervisors, sync licensing, and the rise of MTV.

In Part Four we covered the cassette tape, the CD boom, and how digital audio quietly set the stage for everything that followed.

In Part Five we covered Napster, iTunes, and how the internet dismantled the industry's century old system of control.

In Part Six we covered the rise of the independent artist - cheap recording technology, CD Baby, Myspace, and the connected ecosystem that let artists build careers without label infrastructure

In Part Seven we covered the indie sync boom — prestige television, tech brands, sync agencies, and the workflow problem that led to platforms like DISCO being built to solve it.

Let's pick up where we left off — with a Swedish startup that had looked at everything Napster revealed about what people actually wanted and decided to build it legally.

The pattern breaks - but not for everyone equally

Every major technological shift covered in this series followed a recognizable pattern. A new way to reproduce or distribute music arrives. The industry scrambles. Rights holders lobby, litigate, negotiate. Eventually a new revenue model emerges and the system adapts.

The streaming era broke that pattern - but it broke it differently depending on where you sat in the industry.

For the major labels, streaming didn't break anything. It eventually rescued them. CD sales had been collapsing since the early 2000s and the labels had spent a decade watching revenue erode with no clear answer to the problem. Streaming provided the answer, and the labels made sure they were on the right side of it before the doors opened.

When Spotify needed to license the major label catalogs to make the platform viable, the labels negotiated from a position of genuine leverage - without their music Spotify had nothing. The deals they struck included not just licensing fees but equity stakes in the platform itself. Universal, Sony, and Warner each took ownership positions in Spotify as part of the arrangement. By the time Spotify went public in 2018, those stakes were worth billions. Warner sold its full stake for $504 million. Sony sold half of its holding for $768 million. Universal held on longer.

The majors had essentially traded catalog access for equity and the equity paid off spectacularly - restoring and in some cases exceeding the revenue they'd lost when physical sales collapsed. Streaming solved their financial problem completely.

The question of whether it solved anyone else's is more complicated.

For a deeper look at the equity arrangements, Music Business Worldwide has documented the detail thoroughly - it's worth reading if you want to understand the full picture. Ari's Take also covers the conflict of interest angle plainly and directly.

Spotify and the access problem

Spotify launched in Europe in 2008 and in the United States in 2011, and it solved the access problem that Napster had revealed a decade earlier in a way iTunes never quite managed.

iTunes had given people a legal way to buy individual tracks cheaply and conveniently. But you still had to buy them - each one a transaction, each one a decision. Spotify removed the transaction entirely. A monthly subscription fee, roughly equivalent to the price of one CD, gave you on-demand access to almost every song ever recorded. No downloading, no managing files, no decisions at the point of listening. Just search and play.

The appeal was immediate and obvious. Within a few years Spotify had tens of millions of subscribers and was growing rapidly. Apple, Amazon, and others followed with their own streaming services. The CD effectively collapsed as a commercial format. Digital downloads followed not long after. Streaming became the dominant model for recorded music consumption faster than almost anyone had predicted.

For listeners it was an extraordinary deal. For independent artists the economics landed somewhere quite different from where the majors had found themselves - and understanding why requires looking at how the streaming money actually works.

How the streaming money works

Each stream on a platform like Spotify generates three separate payments — one for the sound recording, paid to the label or distributor; one for the song performance, paid through the PRO to the songwriter and publisher; and one for the song mechanical, paid through the mechanical society to the songwriter and publisher.

The amounts involved are very small. Average per stream payouts vary by platform and by the deal the label or distributor has negotiated, but the recording royalty typically lands somewhere between half a cent and less than a cent per stream. The songwriter's share is a fraction of that. A song needs millions of streams to generate meaningful income for anyone involved in making it.

For major label artists with catalog that generates hundreds of millions of streams, this adds up. For independent artists with smaller audiences it frequently doesn't — at least not to a level that resembles a living income from recorded music alone. The streaming model rewards scale above almost everything else, and scale is unevenly distributed.

The industry negotiated these rates in the early years of streaming from a position of weakness — the alternative was continued piracy, and the labels accepted terms that prioritized getting the model established over ensuring fair long term compensation for creators. Songwriters and publishers fared particularly badly in the initial negotiations, receiving a significantly smaller share of streaming revenue than recording owners. The debate over streaming royalty rates has continued ever since, with periodic adjustments that most songwriters consider insufficient.

Playlisting and the new gatekeepers

In the early years of streaming, the most powerful discovery mechanism on Spotify was editorial playlisting. A small team of Spotify curators controlled access to playlists with millions of followers - Today's Top Hits, New Music Friday, RapCaviar - and placement on those playlists could generate streams at a scale that transformed an artist's career overnight.

In some ways this replicated the old radio model. A small number of gatekeepers with significant reach decided what got heard. The difference was speed and scale - a playlist placement could deliver millions of streams in days rather than the weeks or months a radio campaign required. And unlike radio, streaming data was granular and immediate. You could see exactly how many people had listened, where they were, whether they'd saved the track or skipped it after ten seconds.

For independent artists, editorial playlisting was both an opportunity and a new dependency. Getting on the right playlist required pitching through Spotify for Artists, building relationships with curators, and having the kind of release infrastructure - press, social activity, prior streaming numbers - that demonstrated an artist was worth the playlist's audience. It was a meritocracy of a kind, but one with its own gatekeeping logic.

Labels learned quickly how to optimize for playlisting. Release strategies were built around it. The playlist pitch became as important as the radio plugger had once been.

Independent curators played an equally important role in the early Spotify ecosystem. Individuals building playlists around specific moods, genres, or micro-communities grew significant followings and became meaningful discovery channels for artists with no route into Spotify's editorial team. A well placed track on the right independent playlist could drive real streams and real new listeners. Platforms like SubmitHub formalized the pitch process, giving artists a way to reach curators directly. As with most things in the streaming era, it also became a business - playlist promotion services emerged, some curators began charging for placements, and the line between genuine curation and paid promotion got complicated quickly.

The algorithm takes over

Spotify's editorial playlists are powerful but limited - there are only so many of them and only so many slots. The more significant shift came when algorithmic playlisting became the primary discovery mechanism.

Discover Weekly launched in 2015 and demonstrated what machine learning could do with listening data at scale. A personalized playlist, generated fresh every week for every user, built from an analysis of listening history, saved tracks, and the behavior of listeners with similar taste profiles. No human curator involved. No pitch process. No relationship required.

The algorithm democratizes discovery in one sense - any track in the catalog can theoretically surface to any listener if the data suggests a match. In another sense it creates a new and less transparent gatekeeping system. The signals the algorithm responds to - saves, repeat listens, playlist adds, skip rates - can be gamed, and an industry of playlist promotion services has emerged to do exactly that. Fake streams, bot plays, and playlist placement for pay are real problems that Spotify continues to address.

For artists and labels, understanding the algorithm has become as important as making good music. Release cadence, track length, the first thirty seconds of a song - all are optimization variables. The algorithm doesn't care about albums. It cares about tracks, and specifically about whether listeners engage with them immediately and repeatedly. Songs have gotten shorter. Intros have gotten shorter. The hook moves earlier. The album as an artistic unit, already weakened by iTunes, faces ongoing pressure.

YouTube and Content ID

YouTube sits in a category of its own in the streaming era - simultaneously a platform, a discovery engine, a social network, and the source of one of the most consequential developments in music rights of the last twenty years.

Launched in 2005 and acquired by Google in 2006, YouTube became the world's largest music streaming platform almost by accident. People uploaded music videos, live performances, fan recordings, and full albums. Listeners used it as a free on-demand music service years before Spotify existed. The rights situation was chaotic - enormous quantities of copyrighted music were being streamed on a platform that wasn't paying for it.

Google's answer was Content ID, a system that automatically scanned uploaded videos against a database of registered recordings and compositions. When a match was found, the rights holder had a choice - block the video, track its viewership data, or monetize it by placing advertising against it and taking a share of the revenue.

Content ID changed the relationship between user generated content and music rights in a significant way. Rights holders could now generate revenue from videos they hadn't made and hadn't authorized, simply by registering their content in the system. For major labels with large catalogs this was genuinely significant income. For independent artists without the infrastructure to register properly it was another system that worked better for those who already had resources.

The creator side of this story is worth noting. A generation of YouTubers built audiences and careers around video content that often included music - background tracks, licensed songs, covers. Content ID created constant uncertainty. A video could be demonetized or taken down because of a few seconds of music the creator hadn't realized was copyrighted. This drove significant demand for royalty free and Creative Commons licensed music - an entire sub-industry of music created specifically to be used in content without triggering rights issues.

TikTok and the fifteen second song

TikTok arrived in Western markets around 2018 and introduces a fundamentally different relationship between music and discovery.

Where Spotify's algorithm matches listeners to songs based on listening history, TikTok's algorithm matches content to viewers based on engagement - and music is embedded in that content rather than being the content itself. A song doesn't need to be good for three minutes. It needs to work for fifteen seconds of someone dancing, cooking, or making a joke. The sonic hook, the lyric that lands as a caption, the beat that makes the transition work - these are the unit of value.

The viral TikTok moment is a genuine career event. A track can sit unnoticed on streaming platforms for months and then explode overnight because the right fifteen seconds catches the algorithm's attention and gets amplified across millions of For You pages simultaneously. Labels and artists have started reverse engineering this - writing songs with TikTok moments built in, identifying the clip before the track is finished, seeding content with influencers before release.

The sync implications are also significant. TikTok has negotiated blanket licenses with major labels and publishers to allow their music to be used in user generated content. Independent artists and smaller rights holders have a more complicated relationship with the platform - licensing terms are less favorable and the royalties generated from TikTok use are widely criticized as inadequate relative to the promotional value the platform extracts.

The deeper question TikTok raises is about the relationship between a viral moment and a sustainable career. An artist can have a song reach tens of millions of people on TikTok and convert a small fraction of that into actual fans, streams, and income. Or they can have a moment and nothing more - a song associated with a trend that passes, with no lasting audience to show for it. The platform is extraordinary at creating exposure and inconsistent at converting it into careers.

The distributor explosion and the volume problem

As streaming becomes the dominant model, the infrastructure to get music onto streaming platforms has democratized completely. DistroKid, TuneCore, Amuse, UnitedMasters, LANDR, Ditto, Symphonic — a growing list of distribution services offer to put any artist's music on every major DSP simultaneously, for either a small annual fee or a percentage of royalties.

Some of these services go further. Companies like Amuse offer advances based on streaming history — analyzing an artist's existing streaming data to forecast future royalties and paying a portion upfront. It's a new kind of record deal, data driven rather than relationship driven, with no creative control attached.

The barrier to releasing music professionally has effectively reached zero. Which means the volume of music being released every week has reached levels the industry has never previously encountered. By the early 2020s tens of thousands of tracks were being uploaded to Spotify every single day. The royalty pool — the total revenue Spotify generates from subscriptions and advertising, divided among all streams — gets split more ways every week.

This creates a structural problem for independent artists. The pool isn't growing as fast as the number of tracks competing for a share of it. More music, same money, more division. The per stream rate, already low, is effectively being diluted by volume. An artist releasing music today is competing for listener attention and royalty share against a catalog that dwarfed anything that existed a decade earlier.

The playlist and algorithmic discovery systems that are supposed to surface the best music struggle with this volume. Breaking through requires either algorithmic favor, a viral moment, significant marketing spend, or some combination of all three. Making genuinely good music remains necessary but is increasingly insufficient on its own.

Catalog consolidation

As streaming made royalty income more predictable and forecastable, music catalogs became attractive financial assets in a new way. If a song generates a relatively stable number of streams per year, its future royalty income can be modeled with reasonable confidence — which makes it look, to a certain kind of investor, like a bond.

Private equity firms and major labels began acquiring catalogs at significant valuations through the late 2010s and into the 2020s. Hipgnosis, a publicly listed fund, bought hundreds of catalogs from artists including major names. The major labels made significant acquisitions of their own. The logic was straightforward — streaming had created predictable income streams from proven catalog, and proven catalog was worth paying for.

For artists the decision to sell was personal and financial. A significant upfront payment in exchange for future royalty income — essentially trading an annuity for a lump sum. For some artists, particularly those whose catalogs predated streaming and who had built their careers in an era when royalty income was less predictable, it made sense. For others, selling felt like a loss of something beyond the financial.

The broader implication was that music copyright, once primarily a creative and cultural asset, had become a financial instrument at scale. What that means for how music is valued, exploited, and managed over the long term is still playing out.

AI and the next disruption

Artificial intelligence has arrived in the music industry before most people were paying attention and is moving faster than the rights infrastructure can respond to - which, as this series has shown, is a familiar situation.

AI generated music - tracks produced entirely by machine learning models trained on existing recordings and compositions - appears on streaming platforms in significant and growing volumes. The royalty implications are unresolved. If an AI generated track streams a million times, who receives the mechanical and performance royalties? The model that generated it has no legal personhood. The company that built the model may or may not have licensed the music it was trained on. The songwriters and recording artists whose work trained the model almost certainly weren't compensated for that use.

The legal challenges are arriving. Artists and publishers have begun suing AI companies for training on copyrighted material without permission or payment. The outcomes of those cases will shape the rights landscape for years. The core question - whether using copyrighted work to train an AI model constitutes reproduction requiring a license - has no settled answer yet.

What is already clear is that AI generated music is diluting the royalty pool in the same way that distributor proliferation has, but faster and at greater scale. A model that can generate a plausible track in seconds has no recording costs, no artist advance, no studio time to recoup. It can flood platforms with content at a volume no human artist can match. The industry is at another inflection point - familiar territory, as this series has demonstrated, but with stakes that feel higher than most previous ones.

Where things stand

The streaming era solved the access problem that Napster revealed and iTunes only partially addressed. It gave listeners something genuinely extraordinary — the recorded history of music, on demand, for a monthly fee. And it created a set of economic and structural problems for artists and rights holders that are still being worked through.

The algorithm has replaced the taste driven gatekeeper with an engagement driven one. TikTok compresses the unit of musical value to fifteen seconds. Distributor proliferation floods the market with more music than any discovery system can meaningfully surface. Catalog consolidation has financialized an asset class that was previously understood primarily in creative terms. AI has arrived before anyone has figured out the rights framework to govern it.

And underneath all of it, the per stream royalty rate remains low enough that recorded music income alone is insufficient for most artists to build a sustainable career on.

In Part 9 we look at how artists are navigating this landscape - the tools available, the strategies that work, the communities being built, and why vinyl sales keep going up while streaming numbers keep climbing.

‍