Yesterday, Google announced its answer to Amazon’s Echo home speaker product, which will be known simply as Home. But it also announced what it called “the Google assistant”, which will operate both on the Home device and in a variety of other products. These announcements, and some recent news from Amazon, highlight the fact that these devices, while important, are merely examples of the endpoints that will be part of a broader picture and not the endgame in and of themselves. That understanding is important for seeing these announcements in the proper context and figuring out where this technology goes from here.
Home is a new endpoint for old services
The key thing to understand about both Home and the Google assistant is that they’re new instantiations of old tricks to a great extent. One of the big challenges of Google’s efforts in this area to date is that they’ve been disjointed and hard to refer to in a holistic fashion. Apple has Siri, Microsoft has Cortana, Amazon has Alexa, but Google has only had Google Now, Google voice search, and the “OK Google” function on certain devices along with plain old Google search on the web or inside of apps or widgets.
One of these things, to put it simply, was not like the others. What’s good about the Google assistant branding (though not necessarily the strange use of a lowercase “a”) is it starts to put a name to this collection of functionalities. That, in turn, should allow users to begin to grasp that these things are part of a coherent whole and not just islands of functions floating in a Google sea.
What Home does is give this new assistant a physical embodiment. That should allow users to more easily grasp the concept of what the assistant does and what functionality lives within it. What’s strange, then, is it appears the assistant will take its first bow as part of the Allo messaging app, also announced at I/O, and not within the Home device. Chances are that’s simply a matter of timing driven by the relative development cycles of hardware and software. But it means the assistant will be somewhat buried at first and, quite possibly, in a software product few people will use (but that’s a topic for another post).
The key thing is the assistant is separate from the Home device and that’s actually a good thing. Though Home will perhaps be its best example, what makes the assistant powerful is precisely that it doesn’t live in any single device but exists in the cloud and becomes available to users through a variety of devices. That’s important because we won’t be carrying our Google Home devices with us everywhere we go. Rather, we’ll use our smartphones, tablets, smartwatches, computers, and other devices throughout the day and the Google assistant will be most useful as it follows us around, building a profile on us from interacting with it many times during the course of the day.
Amazon’s vision for Alexa is also expansive
In this context, it’s worth thinking about what’s been happening recently with Alexa, too. It started out as functionality within a single product, the Echo, but it’s clear Amazon’s vision for it is more expansive than that. Not only has Amazon recently introduced two other dedicated hardware products that provide Echo-like functionality in the form of the Dot and Tap, but it’s started putting the core smarts into existing devices as well. The Fire TV box and Stick have been getting Alexa features recently and there are rumors a forthcoming Fire tablet might also feature Alexa prominently. In this way, Amazon is ultimately pursuing the same vision as Google — that of a virtual assistant that’s truly virtual, inhabiting all the different devices we interact with throughout the day.
Amazon’s biggest challenge is it hasn’t provided those omnipresent devices to most of its users. Yes, it’s gained a certain amount of market share with its tablets and TV devices but its smartphone effort failed spectacularly. As such, it remains absent on the most omnipresent device of all. It’s likely that a standalone version of the Alexa app for iOS and Android will appear eventually but, just as the Google and Cortana apps for iOS are inherently second-class citizens to Siri, Amazon will likely never match its success in the home in more mobile scenarios. It’s easy to be blinded by Amazon’s success with the Echo but the reality is its broader virtual assistant strategy will remain handicapped until it solves this problem. It’s overcome that problem in part by majoring less on knowing everything about you than on playing nice with the services that already do, whether that’s third party calendars, music apps like Pandora and Spotify, or your smart home gear. But that’s still a step away from knowing you the way a true assistant does. Google has an advantage here and it’s one it would do well to play up.
Apple has the components but not the home device
Apple, of course, is coming at all this from the opposite end of the spectrum from Amazon. It does provide the omnipresent smartphone and to hundreds of millions of people at that. Its virtual assistant, Siri, is on all those devices and more in the form of iPads and it’s recently added Siri to the Apple TV, too. But as long as the Apple TV requires a screen to perform most of its functions, it can’t truly compete with an always-on device in the home. To be sure, Apple’s vision is one of personal devices and one possible solution is that everyone who needs to interact with such an assistant simply uses their own iPhone or iPad. But of course, many people don’t always have these devices within arm’s reach (or within the sound of their voice), and Apple Watches aren’t yet ubiquitous enough to make up the difference. Does Apple, too, need a home speaker device in the vein of Echo and Home to remain competitive and fill in this gap in its coverage? The evidence it does is getting stronger all the time.
User profiles and shared devices
Perhaps the toughest challenge ahead for all three companies is how to manage individual user profiles on what will inherently be shared devices. That applies to the Home and Echo, of course, but also to the Apple TV and, to some extent, to tablets. Google made a point of talking in its I/O keynote about how your Home and the Google assistant would (with your permission, of course) get to know you over time and therefore get better at their jobs. But that raises the question of who “you” is in this context. Is it the person who set the device up? Is it whoever happens to be asking it questions at any given time? Is it some aggregated profile based on all the members of the household? There are lots more questions than answers for me at this point about how Home (and other similar devices) will resolve all these issues over time.
Siri on the Apple TV deals with this by being utterly impersonal – it provides the same answers to everyone rather than attempting to personalize itself based on who’s asking the questions and deliberately knows nothing about the user. But, if Google wants to make its learning a differentiator, it needs to figure out how to tell the difference between users and share only the right information with the right members of the household. TV box and service creators have long wrestled with this issue and user profiles have usually been the answer proposed, although it rarely works all that well in practice. People hate manually switching profiles and so the solution usually has to be smart enough to tell which user it’s engaging with automatically and respond accordingly. That may require voice recognition, some sort of detection of nearby devices, or perhaps a different trigger phrase for each user. I’m not sure any of this is going to be ready by the time Home launches,o but it’s something Google needs to be thinking about. It’s also an area where Amazon falls short with the Echo – it doesn’t do well with multiple calendars, for example, which is odd for a device that’s explicitly designed to support households and not individuals.
More endpoints to come
I started this piece by saying these home devices will be endpoints but not the endgame. Let me return to that. Yes, these are the newest endpoints in this broader mission of providing intelligent assistants to respond to our needs in a variety of situations. But they won’t be the last – smartwatches obviously have a role here too, but then so do cars, future wearables, and many other devices not yet conceived of. What makes these assistants most useful is they will move through our lives with us on our different devices. That in turn will require either a greater willingness by users to commit to single-vendor device portfolios or by vendors to take their applications cross-platform. History suggests Apple will likely try the former route, while Google and Amazon will probably favor the latter. But this will only get more challenging as the variety of devices on which personal assistants live continues to proliferate.
31 thoughts on “Echo and Home are Endpoints, not the Endgame”
That’s an interesting point. I wouldn’t think identifying a user by voice can be that hard, especially for a device smart enough to actually decipher what’s being said and what it implies.
For non-voice interactions, that’s more of an issue. Maybe ID by device (assuming phones are personal), and for tablets by finger size, type/touch/grip pattern, or use the camera ?
Voice authentication is build in within Google Home according to internal memo in fact that one of Google biggest leverage versus amazon Echo
Interesting – we’ve just integrated ArmorVox with Amazon – so I guess they’re equal
Are you part of Amazon in some way?
“To be sure, Apple’s vision is one of personal devices and one possible solution is that everyone who needs to interact with such an assistant simply uses their own iPhone or iPad. But of course, many people don’t always have these devices within arm’s reach (or within the sound of their voice), and Apple Watches aren’t yet ubiquitous enough to make up the difference.”
That’s an odd take on things given the estimates that total Echo sales to date are around three million and the Apple Watch has sold more than that. What difference is there to make up? There are almost certainly more Apple Watches in the wild being used than Amazon Echos. It seems obvious that a device that can move with you is more useful than a device that cannot move with you. I suspect we’ll see an Apple Network of Things which integrates many devices, rather than one fixed device. Of course Google can tackle the problem in a similar manner via Android devices. Amazon’s success in this space may be short-lived, both Google and Apple have a number of advantages over Amazon.
“Amazon’s success in this space may be short-lived.”
I’ve learned not to count Amazon out of ANYTHING.
You could be right of course, but I see the future of this space as a network of things, more than one device, and certainly a device that moves with you. Amazon doesn’t have the necessary pieces in play.
it is not the device that matter, rather the Cloud base UI voice assistant that can follow you everywhere that makes all the difference, an area where Apple seems to have no expertise.
Nonsense. Apple ‘Services’ is huge and growing, their back end offerings are improving all the time. I’d say Google is ahead on the back end, but to say Apple “seems to have no expertise” is melodramatic gibberish.
At least we agree that the digital assistant needs to be mobile, and that means mobile devices.
no it is not
unlike Google and Amazon that use machine learning to power their voice UI
Apple with SIRI is stuck in their reprogram answer and function node which is the reason why it will fail the more complex and conversational things become, you need machine learning tech to create a great assistance
I’ll be sure to pass this along to Apple. I’m sure they had no idea what Amazon and Google were working on, and of course we can be sure Apple couldn’t possibly be working on anything similar, because they don’t know machine learning exists. Ah, but they will as soon as I forward your comment. Oy. You should make buttons “Apple. Sucking at Everything Since 1976”.
Simply because they know something is importance or want to have it in their product doesn’t translate into having the capability and tool necessary to make it , look at our Google struggle with social system or Facebook with search despite having all the Data necessary at their disposal
Unlike Apple the reason Google, Facebook, Amazon can create Artificial intelligence system using machine learning is because of their long experience in data collection and development of tools to try to understand it as a mean to provide great service to the current user, for them this is simply an evolution of what they do best already You do not become good at this kind of thing just like that because suddenly it’s important.
You don’t create artificial intelligence the way you create an IPhone, that requires a totally different culture and area of expertise, huge amount of data and tools to create a system that can understand and provide context for those data as a mean to create great services that solve real problem.
Amazon has one key advantage: a clear use case of ordering stuff off Amazon. I have a running list of stuff I should order next time around, maybe shouting it at Alexa instead of jotting it down in gKeep then copying it over to a wish list or order would be smoother.
Other assistant have a much more difficult task, either dubious gains or more complex queries.
Yes, this is obvious, I expect Echo will do fine within the Amazon ecosystem, and that could be enough for it to be successful for Amazon. But I do not expect Echo to succeed and scale in the way that Apple and Google will with multiple (and mobile) devices. Echo has been widely available for almost a year now and the sales estimates are at three million. I think an Echo-like device is the wrong approach. I thought that from the very first ad I saw for it. I need a device/digital assistant that moves with me. Think Star Trek communicator badge. Of course that doesn’t mean something like Echo can’t be good at a limited number of jobs-to-be-done (such as a cheap device that makes it easier to order from Amazon), but if it isn’t mobile I don’t think we’ll see huge sales numbers.
I think a strategic tie-in for Alexa could be the smart home. Apple isn’t very good at cooperating, Google is under suspicion of sucking the profits out of their ecosystems, that leaves Amazon as a possible candidate to market and curate a smart home ecosystem of multiple OEMs , with Alexa as the/a UI ?
As for portability, Android at least provides ways Alexa could be fairly accessible: OKg can launch the Amazon app and its embedded Alexa, Amazon can make an Alexa Launcher similar to Google’s with always-on voice recog, or Amazon could have another go at phones, not a crazy & expensive device this time.
A formal ‘smart home’ is the wrong way to think of this. It’s more about bringing intelligence into the home in a way that is mobile and flexible enough to cover a wide range of jobs-to-be-done. Amazon has a lot of work to do if they want to match what Apple and Google can do. Google’s solution will be more open, and Apple’s will be more closed. Both have pros and cons and both offer value in different ways. There isn’t one right way. Certainly it’s possible for Amazon to put the necessary pieces in play to match Apple and Google, but that’s going to take years. Amazon has to beat Apple on hardware and Google on the back end. That’s a tough ask. And they’re going to have to do a lot better than three million Echos sold.
3 million Echos is a hard number for a dedicated smart assistant. Do we know how many iOS/Android users use smart assistants for not 100% endogenous tasks (such as setting up reminders and sending messages) ?
We’ll have to agree to disagree, I don’t care how you define it, three million sold in a year isn’t good enough. Echo clearly has good use cases among a segment of Amazon customers, but at this point I see no evidence that Amazon can expand beyond that.
What I am still looking for unsuccessfully, is any data that suggests consumers really need digital assistants in the first place. Do consumers really want digital butlers, especially if the butlers can’t do the laundry, clean up the mess in the kids room, make the beds, walk the pet?
What is the usage of Siri, Cortana, Alexa and Google Now? How many use it regularly? What do they use it for?
Unless this data is presented, I remain skeptical of the category as a whole, including most bots. To me, it looks like technology in search of a problem that isn’t there. And I am surprised that most people seem to blindly assume that there is a jobs-to-be-solved here.
There are jobs-to-be-done in this space, but they are more along the lines of a number of small conveniences, and consumers will find value in many different places. A digital assistant is very useful, but it has to move with you. I would expect smartphones and wearables to evolve and dominate this space.
I’m wondering whether I’m being an old codger or having an effective BS-dar these days: wearables, VR, smart homes, smart assistants…
In the present case, I think the gains are fairly small (I actually did schedule a movie outing last week-end, took all of 2 minutes), so I think success hinges on execution, with pertinent & speedy suggestions/answers/actions.
For now, Google Now is extremely adept at suggesting web articles I’ve already read (and I’m using Chrome now, so it really has no excuse). On the other hand, it does have a higher hit rate (20% vs 10% ?) than my RSS feeds which are rife with dupes and stuff I don’t really care about. Also, voice recog is a work in progress, it needs to learn to handle ambient voices and sputtering+mumbling+hesitating.
Label me curious and vaguely awed, but not sanguine.
I’ve been wondering this myself… I work at a company that has thrown themselves fully in to the VR content creation world, and it all just seems so…fragile? I could show you thousands of lines of Skype conversations of the VR evangelists that insist its the future – yet every time they SHOW me what the future could be like, it looks nauseating.
BUT…am I being old? Was I a naysayer around the time of the beginning of the mobile revolution? I honestly can’t remember as I think I was too caught up in how cool it all was at the beginning. Is that VR now?
Amazon Echo is more useful than a Smartwatch itself, you don’t have to use something everyday or all the time for it to be essential (IE) Digital Map or Google Map
Of course, the problem is that many people still think that Smartwatches are also a technology-looking-for-solution. Sales are still not yet enough to prove that it was a success.
Being more useful than Smartwatches isn’t very reassuring, unless maybe if you have Apple’s marketing prowess.
voice UI and VR will be very cool and natural to you child just as touch screen was for you and not for your father at the beginning,
What makes amazon echo and Google home cool is the fact that it free you from your phone while still doing a lot unlike your smartwach which is basically just another phone on your whist
Keep in mind that 1 year olds touch (and put into their mouths) everything they can get their hands on but still cannot speak. In terms of intuitiveness and effortlessness, touch is hard to beat.
Touch is good bcuz it is the best right now you should see young teen interacting with voice UI they’re the ones you use it the most primarily at home besides voice UI won’t kill touch it will just add another human natural interactions to the mix and that will also include gesture UI such as project Soli as well
“just as touch screen was for you and not for your father at the beginning”
No. Touch is natural and intuitive. The iPad Air we got for my parents (82 and 78) is the first computer they can use well. If voice UI and VR are also natural and intuitive, humans will use them, age won’t matter. It is bias that matters “Oh, iPads aren’t real computers, real computers have a mouse and pointer, I won’t use that”, and all the while my teenagers (and my wife and I) are doing all sorts of real work on our iPads. My parents are old, but have no bias in this regard, so they love the touch UI.
Maybe you need to go back in time to see what people like you used to say about touch screen such as iPod touch before it becomes extremely popular they used to be as dismissive as you are about voice UI
People like me? I’ve always been positive on touch, as well as voice and VR. VR is a ways out and I think AR will be more useful first, but voice is obviously great. I’m a HUGE fan of touch + voice, always have been (I was using voice commands on my Mac in the 1990s). I realize some very dumb pundits were anti-touch UI, especially when the first iPhone came out, but those jerks were more anti-Apple and were dismissing touch because that’s what Apple was doing. Their bias clouded their vision, they weren’t able to see how obviously great touch was.
This is why I read Tech.pinions – to see an objective analysis, not some over-aggressive self-described “expert” banging loudly on a drum to SELL US on a new concept!
Are you listening Ben Bajarin?