Voice Assistants Could Transform Local — But Not Necessarily How You Might Think
Voice search continues to show signs of being a formidable modality for finding things, including local things. 20 percent of mobile searches are voice according to Google, and speech-to-text processing and AI are getting better by the day. Google Assistant is now at 95 percent accuracy.
Like visual search (a different column), Google is putting lots of muscle behind voice in order to counterbalance search volume declines seen in the smartphone era. Apps have displaced browser-based activity like search, so alternatives to typed queries are the name of the game for Google .
But the world of voice search and AI — also known as voice assistant apps — is widely misunderstood. Generalist tech coverage has painted the picture of an opportunity that resides mostly with stationary devices like Amazon Echo. But the real scale will happen elsewhere.
The first place is mobile. The installed base of 20 million voice assistant devices is great, but dwarfed by 3.2 billion global smartphones — again, where voice search increasingly happens. Moreover, smartphones are a with you all day, giving them the edge in what I call “total addressable time.”
There’s also a more commercially expansive set of queries when out of the home. Alexa and Google Home are great for weather updates, egg timers or e-commerce (Amazon’s obvious end-game), but out of home is where 91 percent of retail spending happens. Target is already thinking in these terms.
But because Amazon dominates the home-device market where everyone seems to focus, many have already crowned it the voice-assistant king. Still, in the more scalable smartphone arena, Amazon is ill-positioned without a device nor operating system. In fact, that touch point is dominated by Google.
Beyond device penetration, as I’ve mentioned, the voice assistant wars will be a won on data and AI chops. On those measures, no one is better positioned than Google to answer the broadest range of questions, given its comprehensive knowledge graph and 15-year tenure as the world’s search engine.
The second area where voice search could scale well beyond the in-home market is my new favorite topic, Augmented Reality (AR). AR is often discussed in terms of graphical overlays to the physical world. But there’s an alternate AR modality with lots of potential for local commerce: sound.
We’re talking ambient whispers for proximity-based social alerts (Facebook), local discovery (Yelp), navigation (Waze) or even pertinent details about someone you’re shaking hands with at a conference (LinkedIn). This is what I like to call “Audio AR,” and one of its advantages is discreetness.
The way it will play out is through increasingly sleek wireless ear buds that condition people to leave them in their ears all day. That creates a new channel for ambient audio. From there it’s up to developers and local media companies to develop apps, including commerce-based monetization.
Google is all over this. Its recently-announced Pixel Buds are explicitly for smarter ways to listen to music and summon Google Assistant. But my theory is that it’s a longer-term AR play to push you audible information throughout the day — back to the point about search volume “counterbalance.”
Apple has this vision too with AirPods. Its positioned well given the installed base of iPhones, but the Android Universe (three billion units) eclipses iOS (600 million units). More importantly, the voice assistant game will again be won on AI chops… where Google Assistant blows Siri out of the water.
Back to Amazon, Alexa has captured the media’s attention (and South Park) because it’s novel and quirky. It’s captured the hearts of 15 million consumers because it’s easy. And that’s the lesson: Though mobile has greater capacity for scale, voice search is only successful when the UI is sleek.
That means voice assistant apps have to be easy to summon on the fly. They should walk a fine line between being easy to verbally wake (minimal taps) without killing your battery. Until mobile voice assistant apps can master that, the Echo will continue to get all the attention. And perhaps rightly so.
Mike Boland is Street Fight’s analyst-in-residence and author of the Road Map column. He covers AR & VR as chief analyst of ARtillry, and SF President of the VR/AR Association. He has been an analyst in the local space since 2005, covering mobile, social and emerging tech.