Tim Merel Contributor
Tim Merel is managing director of Digi-Capital.More posts by this contributor:
- Ubiquitous AR to dominate focused VR by 2022
- How ARKit will make more money than ARCore or Camera Effects (for now)
Mobile AR from Apple (ARKit), Google (ARCore) and Facebook (Camera Effects) and computer vision/machine learning (CV/ML) are focusing the minds and wallets of VCs in Silicon Valley, China and beyond. The $2.5 billion invested in AR/VR so far this year ($1 billion in October and November alone) was balanced across AR and VR, but now mobile AR and CV/ML are the new hotness (as VR has cooled).
Digi-Capital advises a broad range of VCs and startup CEOs, so I spoke to a few friends about how the investment market has evolved. You can read what nearly 30 prominent VCs had to say below, but for the TL;DR crowd:
- Mobile AR and CV/ML are at opposite ends of the spectrum — one delivering new UX/UI and the other powering a broad range of new applications (not just mobile AR).
- Mobile AR is very early stage, and could see $50 to $100 million exits in the next 18-24 months. Dominant companies will take time to emerge.
- CV/ML is more advanced, and could see dominant companies in the medium-term.
- It will take time for developers to learn what works and consumers/enterprises to adopt mobile AR at scale (note: >Digi-Capital’s base case is mobile AR revenue won’t really take off until 2019, despite 900 million installed base by Q4 2018).
- VCs are looking for startups to dominate a vertical first, then turn that into a horizontal platform play (note to startup CEOs: VCs will laugh at your “we are a mobile AR platform” deck today).
- VCs are interested in native mobile AR, not ports from other platforms.
- VCs love CV/ML startups with real-world solutions to fundamentally disrupt industries, not research projects.
- VCs are investing in more than 20 different mobile AR and CV/ML sectors, but they’re not the same VCs (obvious note to startup CEOs: don’t spam; understand VCs as individuals and think about how you fit their portfolio — generalist VCs might make only one or two bets in this space).
- VCs themselves could pose a risk, with the potential for overfunding during the earliest stages of mobile AR.
That’s enough from me. Let’s hear from them.
Aydin Senkut, Felicis Ventures
Felicis Ventures Founder/Managing Director Aydin Senkut invests in both reinvention of existing industries and frontier tech opening new markets, and sees opportunities in mobile AR and CV/ML quite differently. “Mobile AR for us is a consumer opportunity, but one where there needs to be breakout hits to popularize the potential of ARKit and ARCore. That dynamic makes it harder to succeed, but the successes will be big when they get here. One of the challenges is avoiding what the major platforms will build themselves.”
“CV/ML has fewer pure platform plays, but many broad use cases as an enhancing technology across industries — it is genuinely disruptive.” Senkut’s approach with Felicis is all about “engineering serendipity” with “diverse, calculated, risk-adjusted bets in markets where there is a lower level of VC competition. We’re not so presumptuous to think that we know what the future looks like, but we’re confident that when it walks through the door we sometimes recognize it.”
Jenny Lee, GGV Capital
GGV Capital> China Partner Jenny Lee thinks in terms of horizontal platforms (e.g. search, messaging, e-commerce), and the verticals that leverage them (e.g. travel, HR, games). In that context, the firm is interested in how data across platforms and verticals is processed by AI/ML systems for both consumers and enterprises. “We think of AR as one way to represent what comes from CV and other AI/ML technologies, so an enabling technology to create something useful.”
With automation a priority for China, the combination of CV and AR offers unique opportunities. “My favorite example is the humble mushroom. With huge variety in China (several poisonous), picking and sorting mushrooms is done by teams with 10 to 20 years’ experience. We’ve seen startups using CV and machine learning for categorization and labeling mushrooms, then using AR to enable less-skilled workers to do the sorting and selection. Eventually combined with robotics, a traditional part of industry becomes more scalable.”
Lee also sees opportunities in the Chinese education market. “Publishers have no relationship with students, as after they buy the book they don’t come back. The addition of AR features using mobile AR gives publishers an ongoing relationship and a means for further monetization.” In other words, turning a business with 100 percent churn into one with lifetime value.
Matt Murphy, Menlo Ventures
Menlo Ventures> Partner Matt Murphy ran the iFund with Apple after the iPhone launched, which gives him a somewhat unique perspective. “The iPhone was a new computing device and represented a fundamental platform change, as it changed how we use phones forever. At the other end of the spectrum have been API-driven changes, such as location-based and health-focused APIs from Apple and others.”
He sees the scale of ARKit and ARCore as somewhere between the two — not as big as a full platform change, but not as small as an API-driven incremental upgrade, either. “The opportunity for developers with ARKit and ARCore is to leverage a market that is already at scale. They don’t have the chicken and egg problem that most new platforms face, with a huge total addressable market at launch.”
Bill Malloy, Sway Ventures
Sway Ventures> Founder/Partner Bill Malloy is passionate about both mobile AR and CV/ML opportunities. “Mobile AR is not VR, because it solves the distribution issue from day one with an installed base in the hundreds of millions. However the opportunities for investors in mobile AR and CV/ML are quite different. Mobile AR could see a bunch of $50 million to $100 million exits in the next couple of years, where CV/ML could drive a new generation of tech giants in the next 3 to 5 years.”
In machine vision/ML, he sees the value in “capturing, cleaning and owning the data at scale, and then broadening out from vertical products into dominant horizontal platforms.” However, “the talent pool in this space is either in universities or working for Apple, Google or Facebook. So for CV startups, the biggest challenge is people.”
Dr. Shahin Farshchi, Lux Capital
Lux Capital> Partner Dr. Shahin Farshchi is looking to invest in “content and experiences which enable something totally new, and which aren’t possible with any other platform. The creative community will come up with breakout apps that users care about to drive mobile AR at scale.” So while he is investing in both underlying technology and enterprise applications, he believes that consumer apps need to popularize mobile AR first to build the market. “What nobody knows is when this will happen and what it will look like. It could be in the next three months, or it could be in the next 10 years.”
Jeff Clavier, Uncork Capital (formerly SoftTech VC)
Uncork Capital> (formerly SoftTech VC) Founder/Managing Partner Jeff Clavier is an investor in hologram company Looking Glass and CV stealth startups. “I’m actively looking for mobile AR infrastructure opportunities, with a bias towards horizontal platforms over vertical products. The immediate scale of the new market is what makes it interesting, ranging from mass-consumer to industrial opportunities like remote assistance.” Clavier is broadly looking for two types of startups, “practical applications (e.g. Focal Systems) and visionary tech where use cases and business models will evolve (e.g. CV/neural network on a chip stealth startup).”
Gene Munster, Loup Ventures
As a long-time Apple analyst, Loup Ventures Founder/Managing Partner Gene Munster has similar views to Digi-Capital on how mobile AR could evolve as a long-term bridge to smart glasses. “We see 2020 as the breakout year for AR, with rear-facing AR depth sensors in iPhones in 2018 and Apple smart glasses in 2019” (note: >Digi-Capital’s more conservative base case is 2019 for iPhone rear-facing AR depth sensors and 2020 for iPhone tethered smart glasses).
Munster co-founded Loup because he sees “AR as the next way that humans interact with machines, with investment opportunities in business apps, games, location-based AR (i.e. AR cloud) and medical sectors.” His investment thesis revolves around both short-term exits as well as long-term platform plays, “because it is such an early-stage market, we’re starting with app companies that can demonstrate traction, and, if they get it right, leverage that success to become dominant platforms.”
Richard Tapalaga, Qualcomm Ventures
Qualcomm Ventures> Director Richard Tapalaga highlights Qualcomm’s commitment to moving the underlying technology and sensors forward to enable the AR ecosystem. “Based on our deep institutional knowledge of mobile/AR/VR hardware and software, we are excited to further the ecosystem and have committed to a smart glasses future with investment such as Magic Leap. Mobile AR and CV/ML are strong interests for us, from seed through late-stage investments. Developers are flocking to mobile AR because it has scale now (unlike VR), we all know what mobile traction looks like, and there is no distribution fragmentation (again, unlike VR). That reduces the barriers to entry for VCs, because in many ways we’re on solid ground.”
Tapalaga is investing in “core technology across cameras, sensors, ML, CV and AI. We’re also focused on mobile AR, particularly in the enterprise markets around business processes, healthcare, B2B and industrial applications. Consumer mobile AR startups doing something genuinely novel — native mobile AR, not just ports from other platforms — are also of interest to us. We prefer platforms built on top of ARKit and ARCore, much in the same way that Unity was built out on top of iOS and Android.” Tapalaga sees overinvestment as a risk, with the potential for “overfunding of startups during the early mobile AR market.”
Nabeel Hyatt, Spark Capital
Spark Capital> Partner Nabeel Hyatt sees “the camera as an enabling technology similar to GPS. It’s very different to innovations like the app store or Facebook, which were all about distribution. The combination of AR, camera and ML on phones represents fundamental innovation for new products and use cases, but it could take two to three years for creators to figure out what really works. If you think of the iPhone launching in 2007 and Uber launching in 2009, that wasn’t the first use case folks thought about.”
“ARKit is the next extension of what we want to do with our phones, putting them in context with the real world. AR will become ubiquitous over time.” Hyatt thinks the broader CV/ML opportunity is at the stage where huge value creation is happening, because “we’re already four years into the cycle. We’re just as passionate about the mobile AR space, but that market’s only a couple of months old.”
Niko Bonatsos, General Catalyst
General Catalyst> Managing Director Niko Bonatsos met with hundreds of VR companies in the last few years, but thinks “the short to mid-term is hard for consumer VR companies because the friction is too high.” Conversely, he is interested in investing in mobile AR startups as he believes “they are part of the future of media investing that will culminate in smart glasses. It opens up the possibility for new use cases that haven’t existed before, and is also a big opportunity for established IP holders, particularly with games spin-offs.”
However, Bonatsos doesn’t see the path forward as smooth just yet. “Apple’s update of the app store to better promote developers is great, but distribution remains a bottleneck. Incumbent mobile players have a huge advantage over mobile AR newcomers, so the platforms need to take discovery further. Also, the first wave of mobile AR apps are largely derivative of related sectors rather than native mobile AR, and there’s also the physiology of UX to figure out. People don’t hold their phones up in front of their faces for long periods of time, so apps with high frequency, short sessions might prove successful.”
Jacob Mullins, Shasta Ventures