Type “metaverse what went wrong” into a search bar and you’ll get the same explanation everywhere: Meta’s earnings, a falling share price, a corporate rebrand quietly shelved. That’s the story, but it’s not the whole one. While Mark Zuckerberg’s version of a persistent 3D internet was struggling to find anyone who actually wanted to live in it, a far less glamorous cousin of the same technology kept doing exactly what it was built to do. No avatars in blazers attending virtual meetings, just mining engineers in Western Australia training on digital twins of real machinery, rehearsing equipment failures that would kill them underground if they got it wrong the first time. Nobody called that the metaverse. Nobody needed to.
what the losses actually add up to
Let’s put a number on the thing everyone’s been dancing around. Meta’s reality labs division burned through $19 billion in a single year, and the next year wasn’t any better, another multi-billion dollar loss with no clear turning point in sight, per TechCrunch’s reporting on the division’s ongoing losses. That’s not a rounding error on an ambitious bet. That’s a company setting fire to the annual GDP of a small nation and calling it research and development.
Compare that to what Gartner’s hype cycle framework predicted years ago: emerging technologies get oversold at launch, crash into a “trough of disillusionment,” and only the ones solving a genuine problem climb back out. The consumer metaverse did exactly the first two steps and then stalled. Meta kept funding the headset business, kept funding Horizon Worlds, kept insisting the audience was coming. It mostly didn’t, and the spending kept climbing anyway.
Here’s the bit that gets lost in the “metaverse is dead” headlines: this isn’t really a story about a failed technology. It’s a story about a failed bet on where the demand would come from. Nobody was refusing to buy the headsets because virtual reality doesn’t work. They were refusing because there was no compelling reason to strap one on to attend a meeting you could take on a laptop in your pyjamas.
That’s the actual shape of what went wrong with the metaverse. Not “VR doesn’t work,” but “consumers didn’t want the specific thing on offer, and the company kept paying to convince them otherwise.” Which makes the next part of this story more interesting, because somewhere else, entirely unbothered by any of this, the same underlying technology was quietly proving its worth.
General information only. This article is for informational and educational purposes. Technology changes rapidly, and details may have changed since publication. This article does not constitute professional technical, security, or financial advice.
why Horizon Worlds and Vision Pro underperformed
Start with the demo. In the demo, Horizon Worlds was a bustling social space, all avatars and shared laughter, the beginning of a new kind of internet. In practice, most of those spaces sat empty, because building a persistent digital society turns out to be a harder problem than shipping a headset. Meta has now burned close to $19 billion in a single year chasing that vision, and the losses haven’t slowed down since.
Apple’s approach was different in every way except the outcome. No shared virtual society, no metaverse branding, just a genuinely well-engineered headset asking you to pay more than most people spend on a laptop to watch films alone in your lounge room. Impressive hardware, thin reason to use it daily. Meta is still spending heavily on the AR/VR division with results that mirror Apple’s problem from the other direction: brilliant engineering chasing a use case nobody quite asked for.
Here’s the pattern in both cases. Consumers were offered a solution before anyone had convincingly identified the problem. Nobody was lying awake wishing they could attend meetings as a legless cartoon, and nobody’s daily routine had a headset-shaped gap in it. The eSafety Commissioner’s research into the metaverse found real safety concerns worth taking seriously, but even that presumed meaningful adoption in the first place, which the numbers never delivered.
This wasn’t a failure of virtual reality as a technology. It was a failure to answer a basic question before spending billions: what is this actually for, and who, specifically, wants it. In Seoul, gaming arcades tried VR headsets around a decade ago and the appeal faded fast once the novelty did. The lesson was available well before anyone built a second internet around it.
where immersive tech kept growing
While Meta was pouring money into headsets nobody wanted for a virtual office nobody asked for, a much less glamorous version of the same technology was quietly doing its job. Walk into an Australian mine site and you will find VR training rigs that put new operators inside a haul truck cab long before they touch the real machine, running through equipment failures and confined-space rescues that would be too dangerous or too expensive to rehearse for real. Mining companies have been investing in digital twins to model equipment wear and simulate emergency scenarios before they happen underground, and the appeal has nothing to do with avatars or virtual concerts. It is risk reduction with a dollar figure attached, and mine safety officers do not need a keynote to tell them it works.
This is the quiet half of the metaverse what went wrong story. Nobody marketed this as a metaverse, which is precisely why it survived the metaverse’s collapse. Digital twins for infrastructure planning, VR-based safety certification, remote inspection tools that let an engineer walk through a virtual replica of a plant they have never physically visited, none of it needed a shared virtual economy or a Meta Quest ecosystem to justify its existence. It needed a training budget and a genuine cost to reduce.
The pattern is not new. Flight simulators did not wait for consumer hype to prove their worth, and neither did surgical training rigs. The industrial use cases were always the sturdier bet, because somebody could point to the exact problem being solved. The consumer metaverse never managed that trick.
the hype cycle and what comes next
Every emerging technology gets shoved onto Gartner’s hype cycle whether it asked for the ride or not, and the metaverse followed the shape almost too perfectly to be useful as a diagnosis. Peak of inflated expectations somewhere around 2021, trough of disillusionment right about now, with Meta’s balance sheet standing in as the cautionary graph everyone points to. That framing is tidy. It’s also missing the point, because the trough only applies to the version that was marketed at consumers.
The honest answer to metaverse what went wrong isn’t “the technology failed”. It’s that one branch of it was built to sell a lifestyle, and the other was built to solve a cost problem, and only one of those survives contact with a budget review. Mining safety training doesn’t need to climb out of a trough because it was never up on the peak in the first place. It just kept doing the unglamorous work of reducing incidents, one simulated shift at a time.
So what comes next probably isn’t a second consumer wave. It’s more quiet expansion of the industrial branch, and fewer people calling it a metaverse at all, because the word did the technology no favours and dropping it might be the smartest rebrand nobody announced.
Closing / key takeaways
The metaverse story was never really one story. The consumer version, chased by billions in Meta’s ongoing VR and AR losses, landed on a public that mostly shrugged. That part of metaverse what went wrong is settled. But the industrial version kept working the whole time, quietly, without needing anyone to believe in it. Digital twins and VR safety training didn’t need a headset in every loungeroom to justify themselves. They needed one working shift at a time.
Key takeaways:
- The consumer metaverse underdelivered on its own terms, badly
- Industrial VR and digital twins grew regardless, driven by results not hype
- The lesson isn’t that the technology failed, it’s that the marketing overreached
- Watch for the word disappearing while the tools stay
Frequently Asked Questions
Did the metaverse actually fail, or just Meta's version of it?
Meta's version failed, and the numbers say so plainly. Reality Labs, the division behind Horizon Worlds, has burned through more than 60 billion US dollars since 2021 with user numbers that never came close to justifying the spend. But "the metaverse" was never one thing, it was a marketing umbrella stretched over VR headsets, digital twins, avatar-based social platforms and industrial training systems that had little in common beyond wearing a headset. The consumer, social version aimed at everyone genuinely flopped. The industrial version, quietly running in mining sites and factory floors, did not. Conflating the two is how you get a headline that's half right and half misleading, which is roughly what most coverage has managed.
Why did the consumer metaverse struggle so badly?
Mostly because it solved a problem nobody had. Horizon Worlds asked people to strap on a headset to attend a meeting or hang out with legless avatars, when a phone call or a group chat did the same job with less friction and no motion sickness. Seoul worked this out with 3D virtual worlds well over a decade ago. Novelty draws a crowd once. Retention needs a genuine reason to come back, and "socialising is better in VR" was never a claim the product could back up. The tech worked reasonably well. The pitch was the problem, it promised a new way of living when most people just wanted their existing life to run a bit more smoothly.
Is any part of the metaverse actually still growing?
Yes, and it's the part almost nobody was calling "the metaverse" while it happened. Digital twins, virtual replicas of real machinery, mine sites and factory lines, are in genuine industrial use, with Australian mining operators among the most committed adopters, running VR safety training and equipment simulations that cut real costs and reduce real injuries. This corner of immersive tech never needed a hype cycle because it was judged on whether it worked, not on how the demo looked at a keynote. It grew steadily through the exact years the consumer metaverse was collapsing, which is a useful reminder that the underlying technology and the marketing built on top of it are not the same story.
Should we expect the metaverse to make a comeback?
Not under that name, and probably not in that form. The 2021 pitch, a single persistent virtual world where work, shopping and socialising all happen through an avatar, asked for a level of behavioural change that consumers simply didn't want, and there's no evidence pending that changes that equation. What will keep expanding is the unglamorous, purpose-built stuff: training simulations, industrial digital twins, specific tools solving specific problems. That's not a comeback so much as an acknowledgement that this technology was doing genuinely useful work the whole time, just never for the audience or the reasons it was originally sold on.
General information only. This article is for informational and educational purposes. Technology changes rapidly, and details may have changed since publication. This article does not constitute professional technical, security, or financial advice.
