Mark Zuckerberg took five years and over 80 billion dollars trying to create the next iteration of the internet. On March 17, 2026, Meta announced what analysts, investors, and millions of users had long suspected, that the metaverse experiment is a failure. The company had announced that it would close Horizon Worlds on all VR headsets by June 15, 2026, yanking the plug on the flagship product of its most costly strategic bet in corporation history.
The closure of Horizon Worlds does not merely mark the end of a product. It marks the end of an era, a five-year, headline-dominating chapter in which the world’s most powerful social media company wagered its identity, its name, and tens of billions of dollars on a virtual world that never arrived.

On October 28, 2021, Mark Zuckerberg appeared on a virtual stage and declared that Facebook, the social network with over three billion users, would change its name to Meta Platforms henceforth. The name change was not cosmetic. It served as a statement of single-mindedness: Zuckerberg wanted to create the metaverse, a complete immersion of a digital universe where people could live, work, socialize, and transact using virtual reality headsets.
“The metaverse is the next chapter for the internet,” Zuckerberg declared, framing the vision as the successor to the smartphone era. He projected an interconnected digital world where physical distance would be irrelevant, where avatars would replace faces, virtual offices would replace conference rooms, and Meta would own the infrastructure of human connection for the next generation.
The pitch was audacious, even by Silicon Valley standards. Zuckerberg had been preparing for this moment since Meta’s 2014 acquisition of Oculus, the pioneer VR headset manufacturer, for $2 billion. The pandemic, which forced billions of people into remote work and digital social lives, seemed to confirm that the world was ready. Inside Meta, the vision was consuming. The company reorganized itself around it, renamed itself for it, and committed capital to it on a scale that rivaled the GDP of small nations.
From the first day Reality Labs, Meta’s metaverse and VR division, was reported as a separate financial segment in 2020, it has never turned a profit. What began as a multi-billion-dollar annual loss has grown, year after year, into a financial hemorrhage of historic proportions.
| Year | Reality Labs Operating Loss |
| 2019 | ~$4.5 billion |
| 2020 | ~$6.6 billion |
| 2021 | ~$10.2 billion |
| 2022 | ~$13.7 billion |
| 2023 | ~$16.1 billion |
| 2024 | $17.7 billion |
| 2025 | $19.1 billion |
| Total (2019–2025) | ~$80–88 billion |
The operating loss in Reality Labs is expected to amount to 6.02 billion in the last quarter of 2025 alone, compared to 955 million in revenues. In the entire year 2025, the division only earned a total of two point two billion in total revenue as compared to 19 point one billion in losses, which made the business model, had there been any at all, indefensible to the shareholders.
TechCrunch summarized the scale in stark terms: “You’d have to spend $1 million per day for 200 years to spend that much money”.
At the center of Meta’s metaverse ambitions sat Horizon Worlds, a social virtual reality platform launched in late 2021. It was positioned as the digital living room for the 21st century, a place where millions would gather as custom avatars, attend virtual concerts, build digital businesses, and form communities unconstrained by geography.
The reality fell dramatically short of the pitch.
The day Horizon Worlds was launched, the platform did not succeed in keeping users. Its avatars, notoriously modeled with no legs, were the target of much mockery on social media, and a viral image of the blocky avatar of Zuckerberg in the platform itself became the reflection of the divide between the ambition of promotion and the reality of the product. It appealed mainly to young users who were doing what one description termed as chaotic actions, such as throwing digital doughnuts, the last thing that the aspiration professional and social network Meta had imagined.
Despite high-profile partnerships with brands and artists, including virtual concerts hosted by Imagine Dragons and Coldplay, Horizon Worlds never achieved mainstream relevance. The platform was consistently outperformed by VRChat, a social VR experience built by a fraction of the team and budget. By the time Meta announced the shutdown in March 2026, industry observers noted that the user community’s reaction on forums like Reddit was predominantly positive, relief, not mourning.
A single catastrophic decision did not cause the collapse of Meta metaverse. It was the accumulated result of structural problems that compounded year after year.
The metaverse needed a VR headset, a costly, bulky gadget that needs full concentration and space on the part of the user. The Quest headset was technically successful, but failed to penetrate the mass market. During the first half of 2025, Meta only shipped about 1 million Quests, which is a 16 percent drop compared to the previous year. The market of VR headsets shrank in several consecutive years, which proved what a great number of analysts have long been saying: people were not ready to alter their everyday life habits to enter a digital realm.
A Yahoo Finance investigation, based on interviews with over a dozen former senior Reality Labs workers, discovered that there was a split in organizational disorder. Divisions such as Instagram would often lose their managers to head VR teams who would have no relatable experience, which created a culture of misaligned priorities. One of the former employees of Reality Labs, Vasuman Moza, provided a comprehensive report that accused the middle management of the systematic inability to match products to real application patterns. Moza says that leadership killed helpful tools and put teams in pursuit of ideas that no one actually desired.
The timing of the metaverse was not any better. This became the case right as Meta was pouring its utmost capital into virtual reality when ChatGPT was released in November 2022, sparking a generative AI revolution that shifted the world’s investment, media coverage, and consumer fantasies. Artificial intelligence provided concrete, direct value, the capacity to write, create, analyze, and automate, whereas the metaverse required users to spend money on a future that was years away. This comparison was devastating, and the market preferred the present to the promise.
Finally, Horizon Worlds never provided the answer to the initial question any consumer technology should provide: Why should I use it today? The platform had no killer application, no experience that was compelling enough to make it worth the price of hardware, the hassle of installation, or the change of behavior that users would have to make. Other competing platforms, such as Roblox and Fortnite,e showed that immersive digital worlds could appeal to hundreds of millions of users without VR headsets, merely by being fun, accessible, and more truly social.
The financial losses translated directly into consequences for thousands of employees. In January 2026, Meta laid off more than 1,000 workers from its Reality Labs division, approximately 10% of the VR workforce, as part of a sweeping restructuring designed to redirect resources toward artificial intelligence. Multiple internal VR game studios were permanently shuttered, ending years of development work and eliminating hundreds of specialist roles.
Meta CTO Andrew Bosworth acknowledged to reporters that the VR market was “expanding at a slower pace than anticipated by executives”, a measured corporate admission that the original timeline and scale of the metaverse vision had proven fundamentally miscalibrated.
The most notable aspect of the retreat of Meta is the manner in which the firm has redefined itself. By the end of 2025, the term “metaverse” appeared in no fewer than three of Zuckerberg’s own public remarks, and even in earnings calls, where its total omission was noted by analysts as an intentional act of strategic abdication.
Meta is now committing capital to artificial intelligence on a scale that dwarfs even its metaverse spending. The company allocated $60–65 billion in capital expenditure for 2025, with projections exceeding $100 billion for 2026, directed primarily at data centers, AI chips, and generative AI model development. The financial case for this pivot is unambiguous: in the third quarter of 2025, Meta overall revenue reached $51 billion, up 26% year-over-year, powered by AI-enhanced advertising systems that have dramatically improved ad targeting across Facebook, Instagram, and Reels.
Meta’s stock rose 3.4–3.5% on the day the 30% Reality Labs budget cut was announced, offering a clear verdict from investors on the wisdom of stepping away from the metaverse.
Not all of Reality Labs has been written off. The Ray-Ban Meta smart glasses, developed in collaboration with EssilorLuxottica, have emerged as a genuine consumer success, with sales surging more than threefold year-over-year in the first half of 2025. Where the metaverse failed by asking users to enter a separate virtual world, the smart glasses succeed by integrating AI capabilities into the physical world through a form factor, sunglasses, that consumers already wear daily. The launch of Oakley Meta smart glasses in mid-2025 extended the product line and cemented wearables as the credible future of Meta’s hardware ambitions.
The story of Meta metaverse carries lessons that extend well beyond one company balance sheet.
Vision without product-market fit is not a strategy. Meta declared the metaverse the future of the internet before a single mass-market product demonstrated consumer demand at scale. Narrative momentum drove capital allocation decisions that should have been driven by user evidence.
Capital cannot substitute for organizational quality. Meta had virtually unlimited financial resources, and still failed to deliver a compelling product. Internal dysfunction, unqualified leadership, misaligned incentives, and a culture disconnected from actual user behavior have consumed billions without coherent output.
Adjacent technologies can reframe an entire industry. The metaverse never had the opportunity to mature because generative AI arrived and redefined what “the future of computing” meant for both consumers and enterprises. Technology bets do not exist in isolation; they compete for relevance in a constantly shifting landscape.
Hardware adoption timelines are notoriously difficult to predict. Every technology that has required a new hardware category, from 3D televisions to Google Glas,- has demonstrated that consumer behavioral change moves far slower than Silicon Valley optimism.
However, whatever the ultimate verdict of history may be, the metaverse chapter of Meta is one of the iconic corporate narratives of the 2020s. A firm with a very extraordinary financial and technical power decided to gamble its identity on a unique technological vision, and found that a single vision would not suffice at the cost of over 80 billion dollars.
Mark Zuckerberg himself offered a subdued assessment on Meta most recent earnings call. “I anticipate that the losses for Reality Labs this year will be comparable to last year figures, and I believe we are nearing the peak as we begin to systematically lessen our losses in the future,” he told analysts. Whether that optimism is well-founded, or whether it echoes the same confident projections that defined the original metaverse pitch, is a question investors and industry observers will be watching closely throughout 2026 and beyond.

Hassan Tahir wrote this article, drawing on his experience to clarify WordPress concepts and enhance developer understanding. Through his work, he aims to help both beginners and professionals refine their skills and tackle WordPress projects with greater confidence.