In the three months since I penned my initial blog post on “Ubiquity of Voice and the Venture Opportunity,” I’ve been on a deep dive talking with a wide range of people exploring the “Personal Voice Computing” theme, all the way from entrepreneurs building unique applications for consumers to execs at the large platform companies like Amazon. In that short amount of time the market has developed further (see release of Google Home – we now have both that device and an Echo in the NextView Ventures’ offices) and I’ve become more convinced about the coming wave of awesome Voice UI-enabled applications and supporting technologies.
A connected microphone in every room in your house or one perpetually worn in your ear creates always-present voice-directed ubiquitous computing layer. Voice-directed interactions pairs persistent computing access with humanity’s most natural user interface. The result is a new paradigm of Personal Voice Computing. Every major platform company is developing a pervasiveness expansion strategy: Amazon is subsidizing dots to make them cheap and making Alexa extensible to other microphoned devices; Apple AirPods will miniaturize in coming product generations right into your ear.
The forthcoming mass-adoption of voice devices creates the opportunity for three levels of emerging Personal Voice Computing applications*:
- Mobile replacement applications. These are apps which are functional equivalent translations from the previous platform. For example, you can now order a pizza via voice. Arguably, it’s probably even easier to do so than keying it in on the Domino’s mobile app. Similarly, you can summon an Uber. Soon many popular mobile apps which make sense to conduct with a voice conversation will be available. And in a number of cases, the uses can be even be better than with a visual UI because it’s allows for “deep-linked” access. For example, you can just say “Alexa, play country music” as opposed to loading up Spotify app, typing in “country” into search bar, and clicking on a playlist.
- Applications leveraging new distribution. These are a set of voice applications that, because of distribution advantages, go beyond merely replicating what’s available on mobile. First, pervasiveness distribution of input devices beyond typical mobile settings empowers the usage of new applications (a lot of the home automation apps like opening your garage door or watering your lawn from this ubiquity). Or a specific use-case verticalized approach can put voice computing in new places creating new app opportunities (like hotel rooms whereas previously there was friction in downloading apps). Moreover, though it remains to be seen, but potentially most transformative distribution angle is if there will be a moment in time where the platform-shift adoption allows for new apps to ride the distribution of the platform itself (like many games did on top of Facebook when it launched its app Platform). We haven’t yet seen an Alexa Skill go viral because there isn’t current embedded social functionality, but it’s early.
- Fundamentally new voice “native” application sets. These apps are the most exciting, but admittedly the most difficult to describe and predict. The analogy here is that the mobile platform enabled a service like Uber by providing a portable connected computing layer between ride and passenger that wasn’t structurally available on the desktop web. What are those applications which ubiquitous voice-activated computing fundamentally enables which just wasn’t possible before? The savviest of entrepreneurs will answer that question. A few interesting (though perhaps not revolution) examples I’ve seen are the interactive voice gaming apps like The Wayne Investigation, which defy current mobile gaming archetypes.
For the startups, although I’ve met with examples attempting to do so, I think it’s an extremely tough challenge to rival the resources of Google, Amazon, etc. in building a combination device+computing layer from scratch. Instead, one viable alternative approach for startups is to either create a Level II or Level III application above. It is these application types which potentially warrant an entirely new company to take advantage of the platform-shift opportunity, rather than a product extension of an existing company. The challenge is to think beyond the obvious Level I one-for-one replacements to uniquely harness the distribution and the unique attributes of the voice-as-UI platform itself. Further, what native voice applications will harness similar advantageous of previous’ platform winners, like benefiting from network effects? The primary risk here, like all applications, especially consumer ones, hit-driven is the power-curve of adoption.
The other approach for startups is to create an enabling-layer technologies, whether it be a monetization platform, analytics, or developer/production toolsets. The risks in this category are around timing of the application ecosystem, and how much the large platform players encroach into this realm (e.g. surely Google will have something to say about an ad-driven recommendation layer on voice).
I agree with the predictions that “2017 is the year where voice platforms are opening up to mainstream consumers.” This adoption inflection-point will be the firing gun setting off the real startup race for Personal Voice Computing.
* Thanks to my friend Andrew Lau who meaningfully influenced my thinking on characterizing different application classes emerging from platform shifts.