Hugo Barra, Meta’s former Head of VR, offered some unique insight into the XR industry recently with an extensive blogpost that centers around Apple Vision Pro. Barra warns that, like the company’s first standalone headset Oculus Go, the novelty around casual content consumption will probably fade fairly quickly.

Looking back at his time at Meta (then Facebook), Barra notes that Oculus Go was “the biggest product failure” he’d ever been attached to, stating that although casual content consumption was the headset’s raison d’être, the hype wore off pretty quickly.

Here’s Barra’s appraisal of the situation:

Watching TV/movies in virtual reality seemed like such an incredibly compelling idea that we (the Oculus team at Meta/Facebook) built an entire product around that idea — Oculus Go.

Launched in 2018, Oculus Go was the biggest product failure I’ve ever been associated with for the simple reason that it had extremely low retention despite strong partnerships with Netflix and YouTube.

Most users who bought Oculus Go completely abandoned the headset after a few weeks. The full story is much more nuanced (including the fact that the Oculus Go failure got us on the path to Oculus Quest very quickly), but it taught us an important lesson.

Oculus Go | Photo by Road to VR

Barra notes that poor retention for Oculus Go had to do with a few common factors, including user comfort, friction in starting a session when not already wearing the headset, and the social isolation of watching content alone—all of which is true for Vision Pro as well.

Barra concludes that, at least as far as Oculus Go went, traditional media consumption was “not a core ‘daily driver’ pillar but more an ancillary use case that adds some value to other core pillars (such as productivity or gaming).”

Granted Barra says Vision Pro brings more to the table thanks to its better displays than previous VR headsets, which can create “magical movie experiences on occasion,” but those same challenges that Oculus Go contended with basically remain.

Existing VR Games Would Look Great on Vision Pro, But Without Controllers Most Are Stuck

Barra initially moved to Meta (then Facebook) in 2017 from his role as Global VP at the China-based tech giant Xiaomi, becoming head of Oculus and VP of Reality Labs partnerships. Leveraging his experience at Xiaomi, Meta even tapped the Chinese tech giant to manufacture Oculus Go for both the international market and the Chinese domestic market, also branding it under the name ‘Mi VR Standalone’, belying just how big the company expected Go to resonate.

Only a short year after the release of Oculus Go though, the company shifted gears to launch its first room-scale-capable standalone Oculus Quest, nearly abandoning Oculus Go entirely, which in addition to largely relying on Samsung Gear VR apps, omitted motion controllers due to only being tracked in three degrees of freedom.

Then again, that’s where the comparions stop, as Vision Pro has great hand-tracking, millions of apps, and compelling mixed reality passthrough—all of the things Barra says Apple is hoping to use to make Vision Pro “the future of work.”

Newsletter graphic

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. More information.

Well before the first modern XR products hit the market, Scott recognized the potential of the technology and set out to understand and document its growth. He has been professionally reporting on the space for nearly a decade as Editor at Road to VR, authoring more than 3,500 articles on the topic. Scott brings that seasoned insight to his reporting from major industry events across the globe.
  • Cuteulala Artis

    I’m going to have to disagree. The Oculus Go is a brilliant headset because it is extremely light imagine if they had a controllers for this it would be the most comfortable headset on the planet and I would game on it all day without discomfort.

    • NotMikeD

      I think the fact that the Go had such low retention, and was a failing by Barra’s own admission, doesn’t really support this view.

      • ViRGiN

        Cheap, light, throwaway-priced headset could have its usage, but the whole software just has to mature a whole a lot, as well as available passive content like 180 videos.
        If it had a factor of Bigscreen Beyond, it could be something, living in cohabitant with the big old headsets.

      • Christian Schildwaechter

        Barra’s blog is more differentiated. He basically states that wearing (any) HMD is so inconvenient and uncomfortable that people will only accept it for things otherwise unachievable, with media consumption just not being unique and sticky enough. So Go failed because it was a) an HMD and b) focused on non sticky media use. He doubts even the much advanced AVP or any other HMD could succeed as a primarily media device.

        With Quest they therefore focused on games, because the high immersion is only possible with an HMD, so people accept the discomfort. Which is why he criticizes AVP’s lack of VR games or AR apps to provide enough unique benefit to retain user, while Apple positioned AVP mostly as a spatial iPad.

        No HMD fits my face, so I agree that “wearing HMDs sucks” and use them only for short periods and specific experiences, never to watch movies. Most gamers ignore VR, 60% of the Quest owners stopped using it, but many VR fans say then can never go back to flat games. Go had even lower retention, but for some it was the ultimate portable giant screen usable while lying down.

        And 80% of Go’s use was for watching movies. An HMD rejected by most still being absolutely loved by those who found their own unique benefit from it isn’t a contradiction. XR usability is just very personal.

    • VR5

      I just weighed it and it’s 478 grams. So I wouldn’t call it extremely light. Quests are “only” up to 100 grams heavier. But even that does make a difference.

      According to Wikipedia* the Rift was 470 grams but I don’t know if that includes the cable. Anyway, in comparison the Rift really felt light and that was thanks to the strap being rigid and holding it in place. It didn’t feel front heavy.

      (*Mine is stowed away so I can’t weigh it myself right now.)

      • Arno van Wingerde

        OTOH a Q3*S3pro battery headset weighs 1100 grams, and fairly comfortable even if heavy. I really notice it when I shake my head, otherwise it is fine.

        • VR5

          I’m assuming there’s an additional battery in that setup that acts as a counterweight. Front heaviness is the biggest cause of discomfort from weight so yeah, a heavier setup can actually be more comfortable.

  • Octogod

    Passive experiences largely don’t leverage what makes VR so magical.

    Apple Vision Pro seems to be stuck in a 2018 view of the future, where short stereoscopic videos and virtual theaters for Disney films can sell a new product category. We’ve seen that they can’t all the way back with the Go.

    • Dragon Marble

      Here’s the difference through. AVP doesn’t need the cool factor for retention because the screen looks better than your TV. Whether the visuals are better enough for the discomfort will depend on the wearer, but it is a legitimate TV replacement because it clears the quality bar.

      Same for 180 VR videos. It’s finally no longer blurry. And the 3D movie collection is the likker app because how else can you watch them? I am done buying blue ray discs and ripping them.

      • Octogod

        Most people aren’t willing to spend $3,500 on a TV they sometimes use.

        A killer app has to be a unique feature, not one that can be achieved on every device with a screen they own.

        • Dragon Marble

          My point is that people who do spend $3500 won’t let it collect dust because it is objectively better their TVs — which you can’t say for the Go.

          Which other device can you rent and watch a large collection of 3D movies on?

          • ViRGiN

            Reddit is full of people for whom honeymoon is over. It’s not really a matter of price – there is just nothing to do with it. Nobody is getting extra productive wearing this brick, and watching movies alone? Cool for the first few times or watching in bed. But you’re still wearing giant brick, and now made it much more difficult to snack.

            Big screen app was offering 3d movies for ages, and it was a flopped businessplan.

          • Dragon Marble

            Big Screen and 3D TV failed for specific reasons: lack of content, quality or both. AVP is the first ever viable 3D home entertainment system.

          • ViRGiN

            Depends for who. On Quest or Go i could easily just copy tons of different video format files and playback them on the headset. Back in 2018. Or even earlier with Gear VR.

            Obviously the screens and lenses are superior right now, but that is a natural progress.

          • Jonathan Winters III

            The main reason it failed is because consumers who have not experienced it, won’t usually get to experience it, so won’t buy it. That’s a big part of the reason VR/AR/MR remains niche.

          • Christian Schildwaechter

            Barra’s blog implies that Go also failed for one specific reason: it was an HMD, with all HMDs too inconvenient and uncomfortable for most people to voluntarily keep using them beyond an initial phase of excitement. Go’s low end specs played a part, but he doubts that any HMD incl. AVP could survive only on uses case like media consumption that basically work well without a headset.

            Quest therefore focused on gaming, as the extra immersion in VR provided the unique benefit users needed to keep using an HMD. While I agree that wearing HMDs is annoying, I don’t agree that “normal” used like watching movies alone can’t drive XR adaption. Even the Go had a group of hardcore fans, and adding clear lenses, hires streams, HDR, surround sound, 3D, better fit and (someday) lower weight will make HMDs attractive to more and more users.

            IMHO the idea that AVP or any HMD “must” have certain features is flawed. That may be valid for a (mostly) single purpose device like Quest, but not for e.g. phones that today are used for lots of things, from communicating to shooting video, playing games to paying groceries, waking up in the morning or translating posters. It’s more about achieving enough summary benefit to justify using them. And for some just the 3D movie experience will be enough, even if no decent VR game or killer app would ever be released on AVP.

          • Octogod

            You can rent 3D movies on any VR HMD, using apps like Bigscreen and have been able to for about six years. It’s a very niche audience.

          • Dragon Marble

            How many 3D movies are there on Bigscreen? Five or six? There are hundreds on AVP. That’s the difference between a tech demo and the real deal.

          • Octogod

            If you look at the archive org page you can see that there were around one hundred movies. They shuttered the program due to lack of interest, as people would rather just stream in Netflix or Disney+ without 3D.

  • Till Eulenspiegel

    The Oculus GO was my virtual travelling gizmo during the pandemic lockdown. The “Wander” app is the still the killer app for the Quest even today. It is these killer apps that’s selling the system – they should focus more on these unique experiences in their ads.

  • Leisure Suit Barry

    The resolution was way way too low for watching media properly. Tried Netflix a few times on it and was OK for 10 minutes here and there, but that’s about it.

  • VR5

    I’m not even doubting his point but quantifiers like “most” are really vague. We also heard from Carmack that it had much better retention than Gear VR (which basically was the same platform), and even Quest 2 (best sales and retention yet) had only below 7 million MAU on a 20 million install base. >65% is also a lot (if not most) but it is a number we could compare, if we had other numbers.

  • Adam Knotts

    The oculus go was awesome. It was so ahead of anything else at the time. Introduce a lot of people to VR and VR gaming cuz it actually did have some games. There are some experiences that never made to the quest and I have no clue why. Smash hit and groove vr music visualizer were my favorites and never saw a quest release.

    • Jonathan Winters III

      Been chasing the Smash Hit dev for years for a Quest port, but they are dramatically non interested, sadly.

  • Rudl Za Vedno

    I’d consider buying AVP if it offered native steamVR support and controllers just to play flight sims and racing games, but as things stand today I see no reason to upgrade from G2/Q3 combo. 4300€ is way too much for great movie watching experience and some phone apps. Hell you can get 83 inch LG OLED evo C3 4K Smart TV for much less than that.

    • Arno van Wingerde

      I think both Scott and you are somewhat underestimating Apple’s pull. The Go had “experiences” but Apple has a complete ecosystem, unlike the Go, or indeed Meta. Should it really get behind the AVP, expect VR conferences and such before too long… that alone could get the AVP ball running. Get professional 3D CADCAM stuff on board and there is another stream.
      Of course, the current price is ridiculous, but not for professional use. Get an AVnonpro for $1500 while VR apps are popping up might be a great deal… at the same time Meta has time to improve its hardware to a similar level. Ah, we live in interesting times!

      • Jeff

        Your point about ecosystem has merit for sure, but in terms of actual professional utility with CAD and sculpting etc… that’s all dead in the water because of their stubborn desire to be different by launching with no 6DOF tracked controllers which are sorely needed for precise control and manipulation. The same thing that holds back gaming is the same thing that holds back anything professional beyond floating monitor windows.

    • Guest

      Don’t expect Samsung’s offering to be anything substantial. Google has been lacking on XR development and research, as has Samsung. And Samsung wants to be like Apple, which is why they overprice all their offerings. Samsung scrapped their old device to upgrade it with the very same displays that the AVP uses, which costs $700 for a pair.

      I expect the revised Samsung XR headset will cost $2500 at a bare minimum.

      • ViRGiN

        I assume we will be able to spawn app windows like on AVP. At that point, what does avp really do that Samsung could not? Remains to be seen if they will have the hardware to drive Personas-alike.

        • Guest

          In terms of features yeah, I don’t think there is any reason to suspect it will lack anything the AVP will have, except for quality. That is, Android tablet apps are seriously garbage, and it’ll remain to be seen how well their eye and hand tracking will be, if it has real time dynamic occlusion for hands and objects and so on.

          And then of course there is the Persona thing you mentioned. In all the ARCore stuff shown over the gears (not much), they didn’t really have anything like that. Well nevermind, they do have Project Starline, which basically does something loosely similar, but it’s not designed for mobile hardware and is extremely inefficient. I suspect they won’t have an immediate answer to that, which is a big deal because it’s the only thing to address the isolation thing. And for all it’s faults, it’s the only one that now has spatial positioning in passthrough mode.

          • Christian Schildwaechter

            The (very long) blog entry by Barra is actually quite interesting, with a lot of references/sources and a good analysis what Apple can do that others can’t. And just going by basic hardware, there is no mobile SoC that can touch M2 plus R1 performance.

            Most AVP competitors will use XR2+ Gen 2, which by itself isn’t as fast as the M2. It also has to provide all the signal processing tasks AVP does on the custom R1, which is almost as large as the M2 with absolutely insane memory bandwidth for an excessive amount of sensors. R1 handles all of the high quality passthrough by itself, while enabling it on Quest Pro killed performance and battery on Quest Pro and still causes a hit on Quest 3.

            Apple also seems to have a one year exclusive on the Sony microOLED displays. So there are a couple of reasons to suspect that the Samsung HMD will lack things the AVP has. How this impacts the experience remains to be seen, as Apple massively over-provided to be on the safe side, increasing cost, complexity, weight and power draw (up to 40W!). A successor will drop hardware features that proved to be redundant or better/cheaper solved in software.

  • XRC

    Thoroughly enjoyed using Oculus Go at raindance immersive festivals. Saw multiple units deployed at different VR festivals and public marketing events/demos.

    super easy to use, good visuals (compared to contemporary headsets), decent sound with audio headphones, light enough to be comfortable for an hour or two; no need to burn your smartphone battery unlike gearVR or daydream.

  • Christian Schildwaechter

    Some of that is quite off. At 2018 OC5 John Carmack explained Go was a sideproject halfways through Quest’s development. The press compared the image from early Quest prototypes (Santa Cruz, 2016, Rift CV1 with a mobile compute unit) to PS1 or SNES’ Super FX chip, clearly not ready yet. Back then GearVR sold in huge numbers with horrible retention, and Go was envisioned as a quick, low effort intermediate improvement. (“We thought it’ll be like the Gear VR and it’ll be easy, […] But it turned out to be really hard.”)

    Go was announced 2017-10 and shipped 2018-05 as their first cheap mobile HMD for an unknown target group (“We did not have a clear strategy around the Go launch”). It nonetheless found a (virtual cinema) audience, and, with much lower friction than GearVR, was a lot sticker (“The Go is retaining as much as the Rift.”).

    So on the one hand the article cites Barra, at Facebook for nine months when Go was announced, declaring its media focus the cause for bad retention, only corrected by instead creating Quest, a 6DoF mobile HMD with hand tracking useful for regular VR content. Bad news for the media focused AVP.

    On the other hand Carmack, with Oculus since day one, declaring Go a somewhat improvised side project, seeing the same user retention as their PCVR HMD providing a better VR experience than Quest would 18 months later. So an HMD inferior for regular VR content did quite well with a different user group more interested in media consumption on a simple to use HMD optimized for it. Great news for the media focused AVP.

    • perVRt

      Now that Apple and others are taking VR mainstream it seems that he’s feeling guilty for shafting the GO and then going down the dead end path of turning VR into yet another game console. 10 or 20 million gamers are piss in the bucket compared to billions of normal people!

      • Anonymous

        Lmao talk about basic reading comprehension fail.

  • The whole blog post by Barra is incredibly informative. I suggest everyone to grab a cup of coffee and dedicate some time to read it

    • Christian Schildwaechter

      Absolutely worth reading and also well written. Which helps, as it is rather long due to covering a lot of subjects, not because of unnecessary or redundant fluff. So better bring a very large cup or the whole pot.