Meta Didn't Bet on the Metaverse. It Bet Against Two Companies It Couldn't Beat.

On October 28, 2021, a man in a gray t-shirt put on a headset, walked an audience through a cartoon of floating houses and weightless poker games, and changed his company's name from the word everyone in the world recognized to one almost nobody did.¹ He said he'd spend 'many billions of dollars for years to come,' and described a future of a billion users and hundreds of billions in digital commerce - then admitted, in the same breath, that it 'remains a long way off.'⁸ He was right about the billions. As of early 2026, Reality Labs had burned more than $80 billion - and lost more each year than the year before.⁵

The official story is that this was a vision: one founder, one bold thesis about the next computing platform, betting the company on it. That story is too clean. The metaverse bet was not one decision. It was two corporate anxieties wearing the same headset - and that is precisely why no one could ever turn it off.

The bet started seven years before the name did

If the metaverse were a sudden 2021 epiphany, the spending would start in 2021. It didn't. Facebook agreed to buy Oculus VR in March 2014 - announced as 'approximately $2 billion,' though the deal that actually closed in July came in at a precise $2,001,985,000, structured as cash, stock, and an earn-out.³⁴ That is not the move of a company chasing a press cycle. It is the move of a company that had watched the last platform shift - the move to mobile - hand the keys of its own front door to two rivals it could neither buy nor beat. On a phone, Facebook is a guest. Apple and Google own the operating system, the app store, the notifications, and - crucially - the rules on what data an advertiser is allowed to see. A guest can be evicted, throttled, or taxed at the owner's discretion.

Owning the next computing platform solves that problem permanently. If the metaverse becomes the device people live inside, and Meta builds the device, then Meta is no longer the guest. It is the landlord. The Oculus purchase was a cheap option on never being a tenant again - and the metaverse bet is that option, exercised at full price.

The second anxiety arrived with a name on it

Here is the part the visionary story leaves out. The rebrand landed in the same weeks as the Facebook Papers - the cache of internal documents leaked by Frances Haugen that turned Facebook from a company with image problems into a company on trial. Meta's own head of global affairs later admitted they 'considered delaying' the rebrand in light of the leaks and pressed ahead anyway, framing it as 'the right time to come clean about the focus of the company.'⁷ Zuckerberg called the timing 'coincidental.'⁷ Take both statements at face value and you still arrive at something telling: a company under existential reputational pressure chose that exact moment to retire the name everyone was attacking and adopt one that pointed at the future instead of the present.

The bet was real before the scandal. But the scandal gave the bet a job it was never designed for: it became the story Meta could tell about itself that wasn't the story being told for it. That is the overdetermination. A defensive moat against platform lock-in and a pivot away from a credibility crisis happen to point in the same direction - spend enormous sums on a future nobody can yet evaluate - which means the spending answers to two masters and can be killed by neither.

Read that filing sentence again. 'We expect our investments to increase.' That is a company telling its shareholders, in the flat language required by law, that the bleeding is not a phase - it is the plan. And the plan held. The losses didn't plateau and they didn't fall. They climbed, year over year, like a thermostat with the dial removed.

Stack those bars and the total is staggering. Over 21 quarterly reports dating back to 2021, Meta lost $83.5 billion on Reality Labs - an average of roughly $4 billion a quarter, every quarter, for five years.⁶ In the first quarter of 2026 alone the division lost $4.03 billion against $402 million of revenue - it spent ten dollars to bring in one.⁵ A normal company kills a project like that in year two. Meta could not, because killing it would mean admitting which of the two fears it could no longer afford to carry - and it could afford neither admission.

But what if it was simply a great long-term bet?

The fair objection is that this read is too cynical. Amazon ran at a loss for years and built the most important infrastructure of the internet age. Every platform shift looks insane and wasteful right up until it doesn't, and a company with Meta's cash flow can afford to be early. By this view, the $80 billion is not panic - it is patience with a checkbook, and judging it before the device exists is like judging the iPhone from the cost of building the first one.

That defense is genuinely strong, and it deserves a genuine answer. The answer is that a single clear bet behaves differently from an overdetermined one. A clean platform bet has a kill switch: if the milestones miss, you stop. But notice what actually happened - the spending rose every year regardless of how the milestones looked, and Zuckerberg himself said up front the payoff 'remains a long way off' with no date attached.⁸ Spending that accelerates while the goalposts stay invisible is not the signature of a disciplined long-term investment. It is the signature of a commitment that two separate institutional fears are each unwilling to be the one to abandon. The patience-with-a-checkbook story would predict spending that flexes with progress. The overdetermined story predicts spending that only ever goes up. The chart sides with the second story.

Meta set out to stop being a tenant and to stop being a scandal, and it found one purchase that promised to do both: a future so large no current problem could fit inside it. That is the seduction of the overdetermined bet. It always feels like conviction, because two anxieties pushing in the same direction are indistinguishable from one strong belief - until the bill arrives. Eighty billion dollars later, the most honest line in the whole saga is still the one buried in a federal filing: we expect our investments to increase. They did. A bet you cannot turn off was never really one bet. It was every reason you had to keep going, wearing the costume of vision.