“The VR race is back on”

Some interesting thoughts on the current state of XR developments, user retention and AVP versus Quest, from https://x.com/andrewchen/status/1757090367848280075

The VR race is back on, with Apple’s Vision Pro being launched this month, and the Quest outselling the XBox. I’ve invested in a handful of VR/AR/spatial startups, and have some big lessons learned that might be interesting.

The big question:
Is VR/AR/spatial the next big computing platform?

Some thoughts:

  • market size is growing fast but still relatively small
  • retention is just OK
  • VR is stickiest for use cases / demos you might not understand
  • real money can be made
  • getting your app featured still matters a lot
  • VR hero use case + cross-platform is solution to market size

First, it might surprise you to know that the Meta’s Quest is now outselling the latest series of XBox. This is true – there’s nearly 20M units sold of the Quest, mostly the Quest 2, whereas there are ~12M of the latest XBox Series X/S sold. This is a huge deal, and solidifies the Quest as a potential market player, though it’s still smaller than the PS5 (50M units) and Switch (140M units). But hey, at least it’s on the scoreboard.

Of course, this year with the Apple Vision Pro (AVP) there is a significant amount of buzz in the category overall, which should help sales for everyone. This is sort of the promise of the next wave of VR as the next major computing device. Although the total number of units that the AVP is supposed to sell, this year is relatively small, like <500k, I’ve heard from good sources that a second and third generation is already in the works. I think this means that 2025 or 2026 might be the years when the devices get good enough that they can go mainstream particularly at a lower price point.

One of the biggest obstacles for Quest apps at the moment is that retention is good but not great. I’ve asked for retention data from dozens of VR startups over the past few years, and they will often quote numbers as high as 10% or 15% D28 but what the analytics mean by that is 10-15% of users are active the “ENTIRE WEEK STARTING ON D28.” Obviously this is a completely different metric (and a much easier one) than I typically ask on mobile, and it tells you that the best VR apps are retaining a fraction as well as the best mobile apps (where D30 can often be >30%). Furthermore, contrast this to retention on gaming consoles, which are usually much higher retention than PC or mobile. For instance, Fortnite reported a D30 retention of 16% and 21% for PS4 and Switch, respectively, while their D30 for iOS/Android/PC was 6%, 4%, 4%. Unfortunately currently the Quest looks more like iOS/Android/PC rather than console.

The good news on retention: There seem to be several drivers pushing it in the right direction. First, each generation of the VR headsets seem to be improving retention. Some of that is comfort I’m sure, but also battery life and content. Second, I have it from good sources that younger audiences have better retention than older ones. Third the current strategy emphasizes gaming and multiplayer experiences, which plays to the Quest’s strengths. previously there’s been a lot of work on solo beautiful content experiences which make for great demos, but lead to terrible retention over time.

Real money and real traction can be seen on the VR platform today. Some of these insights have led us to invest in startups like Gym Class VR which is a casual social basketball game that’s inked a deal with the NBA for virtual items. Or those who have been following know that games like Gorilla Tag are now reportedly making >$26M/year and reaching >2M MAUs on the platform. As these types of products grow in popularity, and become the initial wedge for VR devices, I think we’ll see that the overall platform becomes more sticky. Those of you who followed computing history, know that the spreadsheet helped sell the personal computer into the workplace. I think it’s true that multiplayer gaming experiences are selling VR into the home, and better hardware, coupled with more of this type of content will ultimately make the platform work.

From a growth perspective one of the problems is that being featured as an app really still matters quite a lot. The Quest ecosystem is sort of similar to the iPhone App Store in 2010 – you one of the powers that be to bless your product, because being featured was a big deal at a time when people were still checking the App Store for new releases all the time. The advantage of this is that teams that are design-oriented and well-connected within the Bay Area ecosystem will find themselves able to take advantage. The downside is that some of the “weird but working” VR apps that are rough but highly engaging (Gorilla Tag is one example) won’t benefit as much as they should.

AVP’s use cases orient around productivity not gaming, and that there’s a case to make that it might work well since it’s backwards compatible with everything we already use computers for day-to-day. On the other hand, our computer set ups for work are so optimized that it’s hard to imagine a head to head replacement anytime soon. But it’s a different approach and we’ll see what happens. (My pet theory is that people will just use the AVP for gaming, initially, and Apple will take years to embrace that)

All of that said, the killer apps are very hard to predict, because even for the iPhone platform it took 3 to 5 years for Uber, WhatsApp, Instagram, etc to emerge – the first gen was Flipboard, Foursquare, and others apps that ended up not working as well. I think we’ll see a similar thing happen in the AVP ecosystem.

All of this is a big step forward on content, because the last few years there was a head fake for Meta Quest. The problem was that VR was touted to be an amazing horizontal computing device and with the amount of dollars being spent, it needed to get to that vision right away. As a result, a lot of apps were developed that went head to head with the laptop-led computing experience we all love, and VR headset simply did not perform well in that comparison. instead, I’m excited to see that Quest’s content strategy feels more like a narrow wedge, expanding into a broader set of apps, rather than trying to boil the ocean right away. Although, perhaps AVP will force it to go horizontal again, which I think might be a mistake. Instead I’d love to see the two platforms take parallel paths, one premium/2D/productivity-led and the other cheaper/3D/gaming-led.

Going back to an earlier point, the VR market is just not that big right now. You can build an VR app with a few million users but it seems hard to build something that can be worth billions. The one hack that seems to get around all of this is to use VR for a hero experience that catches a lot of buzz, and then to add cross platform capabilities on mobile PC etc. We’re seeing this happen with Roblox on the VR platform, of course but apps like VR Chat, Rec Room, and others are able to reach a much larger audience using cross platform.

Ultimately, we are still early on the S-curve towards making VR/AR/spatial an interesting market for venture-backed startups. yet for the first time in years, I’m now seeing top end startup teams pitch us about the Apple Vision Pro and Meta Quest platforms. That’s exciting and I hope that the competition between the two platforms, both benefits new developers, as well as creating more experimentation within the ecosystem.

Many of you might know that a16z invested in Oculus many years back before it was acquired by Meta. but we also continue to be active investing in quite a few seed rounds and Series As.

We’ll be here this year and next, until we make VR work! Excited that folks are building again.