ANALYSIS · An Unfair Advantage Case Study
Real history. Sourced below. No fiction in this one.
The labels did build a legal alternative. They built two, in 2001, and then refused to license their catalogs to each other’s service — so neither had all the music. The failure wasn’t unwillingness to respond. It was that unified distribution required cooperation their structure forbade, which meant they had no right to win in the one market that mattered, and a company with no catalog to protect took it instead.
This piece runs without a narrator from The Unfair Advantage. Most of the series puts a character from the novel in front of a historical case; this one doesn’t, because the story has no single protagonist and the analysis is cleaner in my own voice. The frameworks are the same. The fiction is simply absent.
Sourcing note
This is a factual analysis of the documented history of Napster, the recording industry’s digital services, and the streaming recovery. No dialogue is invented for any real person, and words attributed to real people are their genuine, published statements. Revenue figures come from industry bodies and vary by methodology and territory; they are given as reported. Facts accurate as of 2026; the framework analysis is my own.
- The Suit
- The Services
- The Matrix
- The Aggregator
- The Counterfactual
- What Actually Happened
- Frequently Asked Questions
- About the Stagnation Assassin
The Suit
Napster launched in 1999. The RIAA sued that December, and the suit worked in court and backfired in the market — it made Napster a household name and brought in millions of users who hadn’t heard of it.
The customer diagnosis is the part everyone gets right, so I’ll be brief. People didn’t move to Napster because they were thieves. They moved because the legal option was worse on every dimension that mattered: selection, convenience, price, speed. You could find any song in seconds instead of driving to a store to pay eighteen dollars for a CD containing three tracks you wanted.
And killing Napster didn’t kill file-sharing. It scattered it. The industry then sued Scour, Aimster, AudioGalaxy, Grokster, Kazaa, LimeWire, and onward into BitTorrent — winning most of the cases and losing the decade, because each judgment produced a successor network designed to be harder to sue.
That much is settled and I’m not here to relitigate it. The interesting question sits underneath: the labels also tried to build the legal alternative, and that attempt failed for reasons that had nothing to do with courts.
The Services
In 2001 the majors launched two competing digital services. PressPlay was a Sony and Universal joint venture; MusicNet was backed by the others. They are remembered chiefly by the nicknames critics gave them: MusicNot and PressPause.
They were heavily encumbered by DRM, capped by rules that read like a hostage negotiation — a fixed allowance of “temporary” downloads and streams per month, restrictions on burning or transferring anything — and priced around ten dollars a month against a free competitor.
All of that was bad, and none of it was the fatal part.
The fatal part was that the labels had split into two camps and would not license their catalogs to the other side’s service. So neither product had all the music. You would subscribe and find that half the artists you wanted simply weren’t there.
A music service has exactly one non-negotiable requirement: it has to have the music. Both services failed that requirement, deliberately, as a consequence of who owned what. Everything else — the DRM, the caps, the pricing — was a symptom of the same instinct, which was to protect control rather than to serve a customer.
The Matrix
The Right-to-Win Matrix scores each market segment on three things — current share, right to win, and remaining opportunity — and colors the result. Green means real opportunity plus a genuine right to win: attack. Red means low share and weak right to win: stay out or exit. The discipline that makes it work is recursive segmentation: keep dividing until the segments genuinely differ, because averaging across dissimilar segments produces a number that describes nothing.
Apply it honestly to a major label in 2001 and you get two segments, not one.
Owning and licensing recorded music: green, overwhelmingly. Enormous share, and a right to win that was close to absolute — they held the copyrights, the artist relationships, and the A&R capability. Nobody could take that.
Operating unified consumer distribution: red. And the reason is specific rather than general. Unified distribution requires every major catalog in one place. Assembling that requires the labels to cross-license to a service one of them controls. Which requires each of them to accept a competitor as the gatekeeper to their own customers.
That was never going to happen, and it wasn’t stubbornness. It was structure. Five competing companies, each with fiduciary duty to its own shareholders, each rationally unwilling to hand a rival control of the shelf. Any one of them could have wanted unified distribution sincerely and still been unable to deliver it.
So the labels’ right to win in unified distribution was not weak. It was negative — their own ownership structure actively prevented the thing the segment required. And they attacked it anyway, twice, with PressPlay and MusicNet, which is what the Matrix exists to prevent.
Notice what this reframes. The standard telling says the labels were slow, greedy, or blind. The Matrix says something more useful and less satisfying: they were a set of competitors being asked to solve a problem that only a monopolist or a neutral party could solve, and they were structurally disqualified before anyone made a single decision.
The Aggregator
Which is why the winner had to come from outside, and why its decisive advantage was an absence.
Apple launched the iTunes Store in 2003 at ninety-nine cents a song. The technology wasn’t the edge — the labels had technology. Apple’s edge was that it owned no catalog. It could therefore broker simple, uniform terms with all five majors simultaneously, because no label was being asked to accept a competitor as its distributor. Apple was a party none of them had to fear in the same way.
Run the Matrix on Apple in 2003 and unified music distribution is green, for exactly the reason it was red for every label: it had nothing of its own to protect. That’s the entire asymmetry.
This is worth naming as a general pattern rather than a music-industry curiosity. When an industry is fragmented among competitors who must cooperate to serve a customer, and cannot, the aggregation profit goes to whoever can credibly stand between them. That party is usually an outsider, and it usually looks unthreatening at the time, because its lack of assets in your industry is precisely what qualifies it.
The structural remedy, in framework terms, is Commitment Architecture — a binding mechanism that holds a cooperation commitment when competitive instinct pulls against it. A neutral licensing body, a joint venture with irrevocable cross-licensing terms, a governance structure none of the five could exit unilaterally. Something that operates regardless of how the parties feel in any given quarter. The labels had nothing of the kind, which is why the cooperation lasted exactly as long as it was convenient, which was not long. Apple supplied the missing architecture from outside and charged for it.
The Counterfactual
One correction to the version of this story I’ve heard in a dozen strategy sessions, including some of my own.
The seductive counterfactual is that the labels could have launched, in 1999, a joint venture offering unlimited streaming of everything for $9.99 a month, and owned digital music forever.
That’s anachronistic. Broadband was not ubiquitous; most people were on dial-up or barely past it. There were no smartphones — the iPhone was eight years away — so on-demand music in your pocket had no device to live on. There was no cloud infrastructure to serve it at scale and no unified licensing framework to clear it. Spotify launched in 2008 and reached the United States in 2011, after years of negotiation and equity concessions.
The realistic missed move is narrower and less flattering to hindsight: cooperative, simple, à la carte downloads several years before 2003, instead of DRM-locked subscription silos. And even that required a neutral aggregator and a hardware ecosystem the labels didn’t have.
I raise this because clean counterfactuals are how strategy sessions avoid the real lesson. “They should have built Spotify” lets everyone in the room feel clever. “They were five competitors structurally incapable of cooperating on the modest, possible version” is harder, truer, and describes a situation many of us are actually in.
What Actually Happened
Napster launched in 1999; the RIAA sued that December, and courts forced a shutdown in 2001. File-sharing migrated to Kazaa, LimeWire, and BitTorrent, and litigation continued for years.
The majors launched PressPlay and MusicNet in 2001. Both were hobbled by DRM and usage caps, and critically neither carried the full catalog, because the two camps declined to license to each other. Both failed.
Apple’s iTunes Store launched in 2003 at ninety-nine cents a song, unified the major catalogs, and became the first genuine legal-download success. Spotify launched in 2008 and entered the US market in 2011.
The collapse was real: US recorded-music revenue fell from roughly $14.6 billion in 1999 to about $6.3 billion in 2009, and global revenue bottomed at $13.1 billion in 2014, a twenty-five-year low.
And then it recovered completely. Streaming drove ten consecutive years of growth, taking global recorded-music revenue to $28.6 billion in 2023 — the highest since 1999 — and around $29.6 billion in 2024, with more than 750 million paying subscribers. The industry is now larger than at its CD-era peak.
The majors were not diminished by this. Universal, Sony, and Warner remain dominant and profitable, own the catalogs streaming depends on, hold equity in Spotify, and set licensing terms; label investment in signing and marketing artists reached a reported all-time high of around $8 billion in 2023.
The contested question moved downstream. In 2024 Spotify stopped paying royalties on tracks with fewer than a thousand annual plays, effectively demonetizing well over a hundred million songs, and the payout model tilts steeply toward a relatively small number of superstar tracks. Labels counter with their own data that aggregate payments to artists are at record levels and have more than doubled since 2016. Both claims can hold simultaneously, which is why the argument continues.
Frequently Asked Questions
Did the record labels refuse to build a legal alternative to Napster?
No — they built two. PressPlay, a Sony and Universal joint venture, and MusicNet, backed by the others, both launched in 2001. Both were burdened with DRM and usage caps, and fatally, the labels split into two camps and would not license their catalogs to the rival service, so neither carried all the music. The failure was structural rather than a refusal to act.
Why did Apple succeed where the labels failed?
Because it owned no catalog. Unified distribution required every major label’s music in one place, which required cross-licensing to a service someone else controlled — something competing labels would not do. Apple was a neutral party none of them had to accept as a rival gatekeeper, so it could broker simple uniform terms with all of them and launch the iTunes Store in 2003 at ninety-nine cents a song.
Could the labels have launched streaming in 1999?
No. Broadband was not widespread, smartphones did not exist, there was no cloud infrastructure to serve music at scale, and no unified licensing framework. Spotify launched in 2008 and reached the US in 2011 after years of negotiation. The realistic missed opportunity was cooperative à la carte downloads a few years before 2003 — and even that needed a neutral aggregator and a hardware ecosystem the labels lacked.
Did the music industry recover from the Napster era?
Completely. US revenue fell from about $14.6 billion in 1999 to roughly $6.3 billion in 2009, and global revenue bottomed at $13.1 billion in 2014. Streaming then drove ten consecutive years of growth, reaching $28.6 billion globally in 2023 — the highest since 1999 — and around $29.6 billion in 2024, with over 750 million paying subscribers. The industry is now bigger than at its CD-era peak.
What’s the business lesson from the Napster era?
That an incumbent can genuinely want to respond and still be structurally disqualified. Assess your right to win per segment rather than in aggregate: the labels’ right to win in owning music was overwhelming and their right to win in unified distribution was negative, because it required cooperation their ownership structure forbade. When an industry needs cooperation its structure prevents, the aggregation profit goes to a neutral outsider.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B+ in shareholder value across Berkshire Hathaway, Illinois Tool Works, Whirlpool, and JBT Marel, where he serves as VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he turns proprietary frameworks — including the Right-to-Win Matrix and Commitment Architecture — into field-tested tools for incumbents deciding which segments they can actually win. He is the author of The Unfair Advantage and the Stagnation Assassin series, a Forbes contributor, and an award-winning keynote speaker. Connect with Todd on LinkedIn or bring him to your stage via his speaking page.
Five competitors were asked to solve a problem only a neutral party could solve, and they were disqualified before anyone made a decision. Score your right to win per segment, not in aggregate — and if your industry needs cooperation its structure forbids, the profit from unifying it is already someone else’s.
The discipline of knowing which segments you can actually win is in The Unfair Advantage.