In a recent analysis in Business Insider, Matt Weinberger asserts that Amazon's Alexa voice-based service spells doom to Google. With all due respect, Matt is taking what I consider a very superficial view of the dynamics at play in the nascent voice bot market. Let's break down his rationale and then I'll share my disagreements and perspective:
Amazon's Echo Dominated 2017 CES
"Just ask anyone" he claims, and they'll tell you that Amazon's voice service dominated the 2017 Consumer Electronics Show. He links to several media sites that all agree with the analysis that Amazon's Alexa service did seem to be springing up like a weed throughout CES. This makes sense: up until December 6, 2016 (one month prior to CES), there was only one voice assistant with APIs that original equipment manufactures (OEMs) could utilize to voice-activate their devices, gadgets and appliance. With only one voice assistant to integrate with (and with no cost), it would be shocking if there were any less Alexa integrations with vendors across CES.
There is no doubt that Amazon's pioneering spirit led to the first mainstream voice assistant to enter the market two years ago, giving Amazon a tremendous first-to-market advantage. Further, Amazon's approach to open integration was incredibly strategic and paved the path to the countless integrations announced and demonstrated at CES. However, to claim that Alexa is winning the battle for voice assistant integrations is a bit misleading, because until a month prior to CES, there were no choices available for OEMs - Alexa was the only game in town.
It remains to be seen what happens now that OEMs can choose between two voice assistant platforms. It's true that Alexa had a tremendous head start and it's unlikely that existing deals will change because Alexa is a fantastic service in and of itself. However, Google's Assistant -- Google's competitor to Alexa -- is now an option for product teams moving forward. Savvy product managers will need to weigh many options when deciding which platform to invest in, including install base (Alexa's favor), but also capabilities and co-branding opportunities (Assistant's favor).
It's All About Business Model Alignment
Mr. Weinberger argues in his piece that because Alexa provides a gateway to shopping at Amazon -- and that Alexa's openness and connectivity creates a virtuous cycle for Amazon -- the more devices out there, the easier it is to order things from Amazon. Even Ben Thomson from Stratechery argues that Alexa is Amazon's operating system and is uniquely placed to be the next front gate (and strategic choke hold) on services and capabilities for the modern consumer.
Mr. Weinberger is right -- Alexa is aligned with Amazon's business in a multitude of ways: It drives more Amazon.com shopping. Amazon can start providing value-add services for Alexa, such as the Prime Music service just for Echo owners ($4.99/month). And Amazon can learn from voice requests more about you as a customer, and provide better recommendations for you on the site or in promotional materials. The more Alexa spreads, the more shopping portals and data points Amazon gets for its shopping business. It doesn't hurt that it also gets all of the Alexa OEMs working in Amazon's AWS cloud system either.
However, Mr. Weinberger swings and misses when assessing Google Assistant's alignment with Google's strategic goals. He claims that Google's Assistant (via Google Home and other Assistant-powered devices) will need to rely on annoying ads for it to be beneficial to Google's core ad business. This is short-sighted and misses a huge delta in the difference in underlying architecture between Alexa and Assistant. Google is currently looking to expand beyond being a core advertising company and is looking for new monetization strategies.
In the case of Assistant, Google sees an opportunity to be a true "intelligence agent" for the next generation of computing. The objective? Let Google gain insights about you through your email, calendar, voice requests, phone usage, and browsing behavior so that it can be as smart about who you are as you are. Once that's accomplished, Google can then act as an intelligent intermediary between you and all other services. Whether it's a virtual bartender, a movie theater, a bank, an airline, or whatever consumer service you'd like to use, Google Assistant can connect you to this service and allow you to interact with this service seamlessly thanks to Assistant's machine learning and interaction APIs. Further, because of machine learning, Google Assistant can even recommend services you may not even be aware of that may help make your life easier, more efficient, and ultimately more enjoyable.
When you compare Assistant's objectives to Alexa's objectives, you see how fundamentally different the business alignment is, but you also see how both services are fully aligned with their respective corporate parent's strategy.
How Corporate Strategy Embeds Itself in Alexa vs. Assistant
Alexa and Assistant share some core abilities: answering questions, setting timers, reading the news, checking traffic, playing music, and controlling smarthome devices. In addition to these common capabilities, Alexa makes it easy to buy stuff from Amazon.com, and Assistant makes it easy to play music, videos, shows and movies on Chromecast-enabled devices. But where these two devices part ways is in how they extend to third-party capabilities, and how people can interact with them. This is where it gets interesting because it's here where we see the architecture reflecting Amazon's and Google's different corporate strategies.
Amazon's Alexa extends itself through a concept called "skills." Being available for two years, Alexa has thousands of skills that you can enable via the Alexa app or via voice. Once enabled, you can ask Alexa to use that skill to do something that an independent vendor has developed beyond Alexa's core capabilities. If you add "Skill X" to Alexa, you can then say "Alexa, ask 'Skill X' <something>" and Alexa will reply with whatever the skill provides as output.
Google Assistant extends itself through "services." Being available for about a month at the time of this writing, there are about 30 available now -- it's obviously early days for Assistant services. Unlike Alexa, there is no need to activate a service - they are available to all Assistant users as soon as they are launched. When you say "Hey Google, ask 'Service X' <something>," Assistant actually sends you over to the service to get the info or functionality you're looking for. And you know you're being transferred to another bot because the voice changes. In other words, when not using Assistant's core capabilities, you get an audible cue that you are now in another brand's bot. This provides a bot architecture that is significantly different from Alexa, enabling brands to develop their own bots with their own personalities and capabilities, with the ability to directly interact with the customer. This "handoff" approach may seem less integrated, but I think it lends to a far more extensible solution.
Alexa Skills and Assistant Services In Action
To demonstrate the differences between Alexa skills and Assistant services, let's say I want to check the balance of my Chase bank card.
I say "Alexa, ask Chase Bank for my balance."
Alexa replies "Your Chase Bank balance for Account #xxxxx is $1,434.34"
I say "Hey Google, ask Chase Bank for my balance."
Chase Bank replies "Your balance for Account #xxxxx is $1,434.34"
Notice that in the case of Alexa, Alexa relays the information back to the user, whereas with Assistant, Chase Bank itself replies (with a distinct voice tone that Chase as developed or selected for itself). Therein lies additional opportunities for Chase Bank: thanks to this service architecture, Chase Bank can go well beyond this simple reply. Instead of the above, the reply could just as easily be:
Chase Bank replies "Hi Jon, your balance for Account #xxxxx is $1,434.34"
Chase Bank continues: "Would you like to email you this information for your records?"
Chase Bank continues: "Is there anything else I can help you with?"
This is just one (to be fair, hypothetical) example of Google Assistant's service architecture provides Chase Bank (or any brand) to develop deeper, more personalized, and more useful service experiences with the end user. Compare this to Alexa, which acts as the proxy to all external capabilities. The Assistant's approach to services has implicit value and will be the source of a new revenue stream for Google if Google's Assistant becomes the de facto assistant for people.
Further, Alexa is specifically (and only) a voice-based service. Assistant is a fundamental Google capability that exists not only in Google Home (the voice assistant for the home), but also in Google's Pixel phone, in Android TV devices, and presumably in the future all smartphones and even a web interface.
Google's Assistant Is More Important to Google than Alexa is to Amazon
The irony of this situation is that while Alexa has a head-start and is currently the market leader in voice-assistant technology, Alexa's success isn't nearly as crucial for Amazon's success as Assistant is for Google's success. The reality is that Alexa is just one of many tools Amazon has to get people to buy more from them, whereas Google Assistant is designed to be a core technology platform for Google similar to Google Search (in fact, it's an extension of Google Search). Like Search, Assistant needs to scale in order to get better and to be the "operating system" (or choke point) for third-party services.
Instead of companies buying ads on Google Search, Google imagines a future where ads turn into interactive experiences. This is how Google gets beyond ads -- becoming the interaction point for everyone else's content. Just like they did with Google Search.