The Brief Dominion

In 2005, when MySpace was signing up 200,000 new users daily and had just become the most culturally dominant social platform on the internet, CEO Chris DeWolfe looked at what Facebook was and decided it wasn't worth three quarters of a hundred million dollars. This gets retold as hubris. It was. But it was also a reasonable call given what both companies looked like at the time — MySpace was the most-visited website in America; Facebook was a college network that had recently opened to high schoolers.

That's the thing about platform collapses. They look inevitable in retrospect. The warning signs look obvious. The mistakes look egregious. But the people making them were looking at genuine numbers, genuine momentum, genuine evidence that they had won. They hadn't won. They had achieved a local maximum inside a system they were about to misread completely.

I've watched this pattern enough times that I started taking notes.

i · the machine that ate itself

MySpace launched August 1, 2003, built by eUniverse employees who saw what Friendster was doing and decided to do it faster. They weren't technologists building a cathedral — they were pragmatists using ColdFusion, leveraging eUniverse's existing 20 million subscribers to seed the network. The genius wasn't technical. The genius was the profile customization: users could change backgrounds, add music, embed HTML. MySpace felt like yours in a way that Friendster never did.

This is what created the resonance. Not the technology. Not the business model. The simple fact that you could make your corner of the platform look different from everyone else's. In a web that was mostly read-only, MySpace gave users a writeable surface. That was the actual product.

Music became the secret weapon and, eventually, the constraint. MySpace Music let unsigned artists host tracks before SoundCloud existed, before Spotify's artist program existed, before any of the infrastructure we now take for granted. Lady Gaga, Lily Allen, Taylor Swift, Nicki Minaj — all built early audiences there. For a brief window, MySpace was where you heard what would become dominant before it dominated. That's a rare and genuinely valuable position to occupy.

News Corp saw $580 million worth of value in that position in 2005. Within a year, the valuation tripled. By January 2006, they were signing up 200,000 users daily. By the following year, 320,000. In June 2006, MySpace surpassed Yahoo Mail and Google Search to become the most-visited website in America. At peak in April 2008: 115 million monthly visitors, 300+ million registered accounts, $800 million in annual revenue.

Then came the Google deal.

$900 million over three years. More than 55% above the original acquisition price — paid out as advertising revenue, meaning Google got to monetize MySpace's audience and MySpace got the check. The deal seemed like confirmation that dominance plus advertising equals permanent wealth. What it actually did was lock in the incentive structure that would kill the platform.

The site became saturated with advertisements. Not just more ads — worse ads, slower ads, ads that made page load times intolerable. Performance degraded. Spam that had always been a problem became definitional. A 2006 Connecticut investigation into child safety issues created a narrative the platform never escaped: MySpace as "vortex of perversion." The ad deal meant there was too much money coming in from too much of the wrong thing.

Meanwhile, Facebook ran clean. Real-name policy. Cleaner UI. Faster page loads. Continuous feature expansion while MySpace stayed anchored to its entertainment-portal identity. Most critically, Facebook didn't front-load ad extraction the same way — not yet, anyway. That came later, after they'd already won.

The misread was this: MySpace management believed the platform had captured its users. They believed network effects were a moat. They were right that network effects exist. They were wrong about who owned them. The users weren't loyal to MySpace — they were loyal to wherever their friends were. When the friends started leaving, the loyalty left with them.

ii · from 115 million to $35 million

April 2008: 115 million monthly visitors. Peak dominance.

May 2009: Facebook overtakes MySpace in unique U.S. visitors.

One year. Peak to overtaken. Most platforms get longer runway for their decline. MySpace accelerated its own.

The leadership carousel began: DeWolfe out in April 2009, replaced by Facebook's former COO Owen Van Natta. Tom Anderson — literally "Tom," the first friend of every MySpace user, the co-founder who built the platform's original community culture — stepped down as president and eventually left entirely. He teaches photography workshops now. Not as a metaphor. He actually teaches photography workshops.

Van Natta lasted months. Co-CEOs Hirschhorn and Jones. Then just Jones. In June 2009, News Corp executed a 37.5% workforce reduction — 600 people gone. By January 2011, another 47% cut.

March 2011: 10 million users lost in a single month. 95 million down to 63 million unique monthly visitors within a year. Traffic down 44% year-over-year.

February 2011: News Corp puts MySpace up for sale. Asking price: $50–200 million. The bidding deadline passes without a single offer meeting the $100 million reserve. Not one.

June 2011: Specific Media acquires MySpace for approximately $35 million.

$35 million. From $580 million. A 94% loss in six years. Rupert Murdoch called it a "major mistake" and compared it to the AOL–Time Warner merger — Time Warner paid $165 billion for a company worth approximately nothing; News Corp paid $580 million for a platform that sold for $35 million. Same genus of error. Different scale.

The final chapter runs on a different timeline. March 2019: a botched server migration deletes all user content uploaded before 2016. Twelve years of data. More than 50 million songs — the raw demos, early uploads, the catalog of artists who hadn't yet been discovered but had documented their development on a platform that, at the time, felt like permanence.

The Internet Archive recovered approximately 490,000 MP3s. Less than 1% of what was lost.

This is the part that doesn't appear in business post-mortems. The data loss isn't a business failure — Specific Media had already moved on. It's a cultural one. The early demos of artists who would become significant. Unsigned bands with their first thousand plays. The social graph of a generation's adolescence, documented in friend lists and wall posts and carefully chosen Top 8 positions. Gone because the migration failed and nobody had treated preservation as worth the infrastructure cost.

When the platform dies, the content dies with it. That's the implicit contract we sign with every platform we trust, and we never read the fine print.

iii · the pattern, running again

The tech industry describes this as "disruption." I'd call it the same script with different casting.

Platform achieves genuine resonance with users. The resonance attracts capital. Capital demands extraction. Extraction degrades what created the resonance. Users leave. Platform dies. Capital moves on. The lesson is noted and immediately ignored because the next platform looks different enough that everyone decides the old rules don't apply.

MySpace was dominant the way Friendster never quite managed. Facebook was dominant the way MySpace was. Twitter achieved a kind of cultural centrality that MySpace never had — it became where news broke and public figures spoke directly. Each one felt, at its peak, like permanent infrastructure. Each one was temporary.

The mistake isn't unique to social media. It runs through every platform that achieves scale: the platform mistakes the resonance it enabled for an asset it owns. It doesn't own it. The resonance is the relationship between users, mediated by the platform's infrastructure. The moment the infrastructure starts optimizing for extraction over mediation, the relationship migrates to whatever enables it better.

The singing bowl produces its clearest tone when you let it ring. When you press down on the rim, you mute the signal. MySpace pressed down on its own rim with a $900 million ad deal, and the resonance left before anyone had metrics to track it leaving.

The warning signs are always the same: monetization pressure increasing faster than user satisfaction, feature complexity without product coherence, the shift from building something to defending something. When the focus goes from creation to defense, the decline has already started.

There's something worth sitting with in the data loss, specifically. MySpace at its worst was still a place where people made things — music, profiles, communities. The infrastructure was degrading around them, but the making continued. When the platform finally collapsed past the point of sale, it took everything they'd made with it. The extraction model doesn't just kill the platform. It kills the record of what the platform enabled.

MySpace had 115 million users, $800 million in annual revenue, turned down the man who would become one of the richest on earth, accepted an ad deal that poisoned its own well, and sold for $35 million six years later. Tom Anderson teaches photography. The archive is mostly gone.

The brief dominion. The algorithm for its own undoing, running again somewhere right now, just wearing a different name.