4 rules for starting a venture-backable AR/VR company today
This is effectively pitching advice for AR/VR founders.
While I’m still deciding whether I love or loathe the “VR guy” label I’ve accumulated in VC land — despite Earthling explicitly investing in much more than AR/VR — I routinely have conversations with VCs exploring the space (mostly in the advent of the Apple Vision Pro) and new builders starting companies about the lay of the venture land in AR/VR.
Specifically, when discussing venture-scale opportunities and limitations in the ecosystem I find myself repeating the same 4 talking points— so I figured it made sense to codify them as the (current) 4 rules for starting a venture-scale AR/VR company.
As always, I disclaim that 1) this is solely my opinion and 2) our job is to have opinions but hold them loosely, as all things in emerging tech develop (and change) quickly. Maybe some of these rules age better than others.
4 rules
Content is king.
There is an age-long debate within the VR community diagnosing the perceived “slow” adoption of VR in the mainstream: is the problem the hardware or simply the lack of high-quality content? While consumer VR devices have plenty of room for improvement, we do categorically observe that good content in VR performs very well and drives significant revenues for small teams (e.g. Gorilla Tag). In our view, the pervasive problem in VR is a shortage of high-quality, valuable, and retentive content. As such, the #1 priority of any team building in VR should seek to uncover a novel, killer user experience for the headset.
This infers that VR studios are solving the most present issues in VR (and thus more likely to get funded). On the flip side, this means the market is currently not ripe for VR platforms (e.g. no-code content creation platforms, etc), which theoretically drive more exponential growth potential and value than content. The market, however, is not ripe for this sort of tooling as demonstratively a) there are not that many builders (i.e. building a VR app is nowhere near as prevalent as building a website) and b) there are not enough discovered killer mechanics or concepts that can be platformed out.
Folks insistent on building platforms must lead with content like Epic Games did with Unreal Engine— they created successful games first, then platformed the infrastructure out. Founders building platforms today must be able to demonstrate that someone (including themselves) has been able to use their tooling to build successful content (e.g. ShapesXR is one few examples to have done this). Building tooling for the sake of tooling in a burgeoning market is, frankly, worthless.
All of the above is true for any mode of extended reality: AR, MR, VR.
Foreseeably, the Apple Vision Pro is not a venture-scale market.
Many VCs and builders are rightfully taking a lot of interest in AR/VR through the lens (pun intended) of the Vision Pro, and its long-term impact and market potential. While Apple’s entry, of course, is broadly a market validator, all signs point to Apple’s capture of the AR/VR market to be a slow burn: Apple is rumored to only be shipping ~500k units in 2024, and iterations on the device are not expected to ship until 2027.
Assuming those estimates to be true, Vision Pro’s addressable market is not large enough to accommodate a company founded today that has the potential to grow on the expected VC trajectory. That being said, not every company should be venture-backed. There may be ample opportunities for bootstrapped teams in the early innings of the Vision Pro, which can always become a venture-backable enterprise when the TAM catches up— but at this time, a company exclusively targeting the Vision Pro is nearly impossible to underwrite as a venture capitalist. Enterprise applications might be the only exception.
By all accounts, the Vision Pro is a spectacular device that’s worth building for. For founders that want to build for the Vision Pro, but be venture-backable today, consider the following: 1) Build an app that's also viable on the Quest platform, or 2) build a hero use-case for the Vision Pro, but generate value across the vertical stack of Apple devices (e.g. phone, watch).
Validate your concept before attempting to raise your first tranche of institutional capital.
Across all of venture, valuations appear to be stable but the expectation at each stage, to me, has appeared to moderately increase over the last few months in particular (my observation aligns with what Jason Calacanis tweeted the other day). This is naturally true for VR as well.
However, I will go out on a limb and claim there is an additional dynamic at play in VR funding. Presently, the number of VCs reliably writing checks into pre-seed/seed VR companies is extremely finite. As such, most of these VCs have made a number of bets into any number of commercially viable VR apps in gaming, mental health, fitness, etc. For new founders, this means a) the chances of conflicting with the VCs’ portfolio companies are non-negligible and b) the bar to impress is high (i.e. why is your gaming studio more compelling than the 5 studios Fund X already has in their portfolio?)
While crazy traction at pre-seed is not expected, validation is (sustained traction is effectively the delta between pre-seed and seed). Validation, at this stage, demonstrates bottoms-up demand for your solution in a venture-sized problem space and your approach to going to market or distributing your product. For example, I love studios right now that are able to successfully build off-platform communities (e.g. Discord, TikTok) that can supercharge distribution (thus not being at the mercy of big companies for distribution while also demonstrating gamers are excited about their title).
3D on 2D screens is not AR or VR.
Geo-anchored virtual objects displayed on your phone are not AR. “3D websites” on your laptop browser are not VR. Other than gaming, where we do play in 3D environments (i.e. virtual worlds) through 2D screens (though I believe this will largely shift to VR), placing 3D content onto 2D planes is not impactful as the value proposition or canvas of a spatial computing experience cannot be realized. As such, it is generally a misnomer, for example, to equate “AR TAM” to the number of smartphones.
In the case of mobile “AR”, I refer to these as filters. While theoretically, this may be a proof-of-concept for the eventual advent of outdoor, persistent-wear AR glasses, that reality is also extremely far away (having spent years working on AR glasses at Meta, I still consider such a form factor to be science fiction for consumers). It’s important to recognize phones do not enable “immersive experiences” (otherwise we wouldn’t need HMDs), they simply enable neat overlays on what we see through its camera. Can valuable products be generated through this medium? Sure. Have I (or the market) seen many of them outside of Pokemon Go, which was likely carried by the Pokemon IP? Not really (and that’s probably not a coincidence).
These phone/web-based 3D platforms are also often guilty of using their theoretical larger device TAM to justify the creation of “no-code” platforms (which we covered under Content is King). If you’re going to commit to the path of using a 2D device as an entry point for your 3D experience, once again, it is important to validate whether people want e.g. 3D websites or a bunch of geo-anchored experiences lived through their phone camera. These might make ok businesses— but they do not share the upside of true AR or VR apps, and in my personal opinion often feel more gimmicky and value-generating.