Google and Amazon Escalate Voice ‘Platform Wars’

This post is the latest in our “Beyond the Screen” series. It will be an editorial focus for the month of January, and you can see the rest of the series here. 

Every year at CES, a theme prevails. This year it was the connected home, and a central component of that is voice search, empowered by personal assistant front ends, à la Google Assistant. Google blitzed last year’s CES with Assistant promos everywhere, prompting this year’s AI-fueled connected-home counter attack.

This is materializing into a two-horse race between Google Assistant and Amazon Alexa (Apple tends to stay away from CES and is quiet on Siri integrations generally). In early stages of what’s looking like another platform war, CES saw Assistant and Alexa integrations in everything from cars to toilets. 

But amidst this convention-fueled game of one-upmanship, a few data points emerged that have more lasting local commerce implications. The most notable was that Google Assistant now has a total installed base of one billion devices. This compares with Alexa’s 100 million, a 10x difference.

As platform wars go, the first step is to establish such an installed hardware base. That then attracts developers to build apps and experiences for the platform (a numbers game), which in turn attracts more users. And things ratchet up in that virtuous cycle. So on that measure, Google Assistant has a clear lead. 

Going one level deeper, the majority of Google’s one-billion device figure is smartphones as opposed to its smart speaker, Google Home. Conversely, Alexa’s installed base of 100 million devices is mostly at-home Echo devices (and some appliance integrations). So, as we’ve examined, Assistant is mobile; Alexa is at home.

This delineation was accurately reported in lots of CES coverage over the past week. But one thing we take issue with is the notion that Google’s mobile positioning is a detriment. In lots of coverage, Amazon’s Echo base is presumed to be advantageous because its core function is voice search, as opposed to smartphones’ varied use case. 

There could be some truth to that given that all one billion Google Assistant-compatible devices don’t represent active users, whereas Echo devices are built explicitly for voice. But there are a few signals that indicate the opposite could be true in terms of optimal user touch points for voice and assistant apps.

First, quantitatively speaking, mobile (and Google) could be at an advantage if you look beyond Google Assistant compatibility to the broader use of voice search. As covered in our voice search white paper (download here), 20% of mobile searches happen by voice. That’s about 120 billion annual searches according to our calculations.

Quantitatively, those searches carry higher commercial intent. This traces back to mobile search’s overall intent dynamics. About 30% of mobile searches have intent to see, eat, or buy something locally. If that same on-the-go intent exists in the subset of mobile voice searches (arguably), they’re well positioned.

Let’s compare that to the intent of at-home smart-speakers. A great deal of evidence indicates that there isn’t much actual commerce happening, with only 0.2% of Alexa users reporting buying anything on the device. It’s rather lots of “portal” functions such as weather, news briefs, and entertainment use cases. 

There’s nothing wrong with those use cases, but if we’re looking at these platforms through the lens of monetization potential, or influencing local commerce, Google Assistant likely has the edge. There’s also Google’s inherent advantage in its knowledge graph, which engenders a better overall voice search utility. 

Why is all of this important? Any entity competing for local commerce—publishers, brands, ad-tech players—has a looming platform choice for voice. Like the platform wars between iOS and Android, it’s a matter of deciding where to apply finite resources and development muscle. Maybe the answer is “both” Google and Amazon. But for now, Google appears to have the lead.

Beyond platform choice, it’s a matter of optimizing presence in voice search. Without getting into SEO and listings optimization tactics for voice (a different column), it starts with reverse engineering user intent and choosing the right platform on which people are searching. The answer will be different for every brand or publisher. 

Back to Apple—it’s quantitatively formidable due to default positioning on about a billion global iPhones. But it’s qualitatively lacking in voice and AI chops (just try using Siri for anything), so that needs to improve before it can compete. Either way, the voice platform wars will be a moving target that we’ll continue to track. 

Mike Boland is Street Fight’s lead analyst, author of the bi-weekly Road Map column, and producer of the Heard on the Street podcast. He has been an analyst in the local space since 2005, covering mobile, social, and emerging tech. More biographical information can be seen here.