In a recent tweet, Benedict Evans brings up the point that business models are important for understanding the nascent smart speaker market. This is particularly important given a Reuter’s report that both Amazon and Google likely lost money on these devices during the holiday shopping season.
This brings up the question, what business models are these smart speakers really suited for, and have Google and Amazon exhausted the possibilities? Are there other companies with different business models that might enter the smart speaker market, and be even more aggressive or successful?
Nowadays, if you want to search for perspective, it is a good idea to look to China to see what the Chinese are doing (other East Asian countries are also illuminating). According to a report by Activate, there are quite a few interesting developments. In the chart below, Activate shows that in addition to Baidu and Alibaba which can be considered as counterparts to Google and Amazon respectively, we see the Asian messaging giants – Tencent (WeChat), LINE and Kakao getting into the game. What’s interesting is that these messaging apps are not just for messaging, but are actually portals to a whole variety of services spanning ride-hailing, deliveries, music streaming, digital payments and more. This means that without plugging in third party apps like how Amazon is doing with their “Skills”, WeChat, LINE and Kakao may be able to provide a battery of useful services that could be better integrated into the voice UI. They would also be in a better position to monetise directly from a variety of services compared to Amazon which only monetises through shopping or Google which doesn’t yet have a clear monetisation strategy yet.
Therefore, it seems that the East Asian tech companies have a better business model for Smart Speakers than any of their US-based counterparts. US customers are reported to use Smart Speakers only for very basic tasks, but is it possible that East Asian users will soon adopt more complex use-cases, simply because the better matching business models will encourage the services to be much better, more varied and more integrated. Who knows? If Smart Speakers turn out to be really successful, they might turn out to be the vehicle on which Chinese companies will finally penetrate the US market (although there might be significant national security issues with having Chinese ambient microphones in US households).
Naofumi Kagami 
Pushing in-progress stuff onto the market w/o a clear understanding of how it’ll make money, but, even before that, of how it will be used, smacks of a frantic search for the next Big Thing and fear of getting locked out. . I think we’re looping back to 3 issues:
1- vague tech vs killer app. Having a smart speaker that can do stuff is nice. But does it do anything that my other gizmos either don’t do at all, or do significantly less well ? That’s compounded by the fact that smartspeakers’ UI and capabilities are utterly undiscoverable. I’m curious what use(s) will catch on, aside from playing audio/music and acting as a smart remote for that. I’ve started forcing myself to use voice typing more, as a stepping stone, but anything beyond that is iffy… Cortana for some reason changes “set an alarm in an hour” to “set an alarm at the next hour on the dot”, that’s an inauspicious start ^^
2- Are Assistants the new platforms ? That’s what your Chinese examples point to. I’m curious what the end state will be… Sci-fi spaceships all seem to have an all-knowing AI that’s sometimes sorry it can’t do specific things. It’s usually a lot more conversational, proactive and intrusive that what we’ve got now.
3- There’s a huge training/generational issue. I never could get my mom to voice-dial nor voice-text on her smartphone, but when my 5yo nephew tried out my brother’s xmas GearVR, he kept trying to talk to some bee that was in there (and was surprisingly adept at chasing it around in the VR while NOT running into furniture ^^). I doubt I’ll ever choose a loudspeaker over a screen-equipped computer for anything but the most basic, dead-end stuff (ie, not searches, those always are russian dolls), but maybe I’m a dinosaur ?
LikeLike
I agree that smart speakers are not very useful, at least not yet. What I would argue is that the “killer app” will likely come from east Asia and not from Google/Amazon. It’s very possible that Silicon Valley pundits will totally miss this.
LikeLike
Semi-off-topic: have we agreed on what the killer app for Mobile was, or is “killer app” an outdated concept of when computers where cost-justified investments to do specific tasks, not magical delightful gizmos in charge of making us wonderful people because we’re well worth it ?
LikeLike
I don’t think that there is any agreement. However, the concept of a single app/use-case being a dominant reason for purchase in the early stages of a market, still applies in my opinion. For mobile, that would have been the browser and the “whole” Internet (prior to the explosion of apps, and when Facebook was still HTML5 in mobile).
Once Apple managed to put a more-or-less complete browser onto a phone, they solved the killer app issue. It was easy because all that was needed to be done was to shrink a desktop-quality browser into your pocket. The utility was already defined on desktops and the issue of shrinking was purely technical. For smart speakers, finding the killer app will be much harder because we still don’t know what useful things there are to be done with voice alone.
LikeLike
I”m always uncertain because the subject I’ve got most insights about (Me !) might not be representative. In my case, the killer app that triggered my early adoption of an HTC HD2 was allowing me to consolidate phone + Palm PDA/ebook reader + MP3 player into a single device. And making it worthwhile to carry a much bigger (so, better) and more expensive device because it was still less bulky and less expensive than the 3 it replaced. The rest has proven very secondary, mostly RSS and real-time browsing replacing my morning site hovering + offline synching routine.
So, let’s say the media thing was the early killer app that justified adding smarts to a phone, then web, then apps. Early on the process was painless because it was replacing devices with a better device, then later on always-on web and apps (and ego crutch) took over.
What worries me is that smart speakers, like smart glasses and smart watches, don’t fit that smartphone model. They don’t unequivocally replace and improve on (a) previous device(s), except maybe if all your music listening is Spotify streams, and if you miss FM radio. They don’t enable something that can’t be done any other way, at best they make it easier / hands-free.
LikeLike
I’m not sure if yours was the typical reason why people bought smartphones early on, but nonetheless, you tell me that it was a replacement and improvement on what you already had and I agree that this is the core of the discussion. I agree that smart speakers do not fit this model, and that will make adoption slow to say the least.
One thing that I think would be worthwhile to investigate is the Bluetooth speaker market. What kind of speakers are selling at what kind of volumes? We could also look at the hi-fi speaker market in general to gauge the size of the high-end market as well. I have a hunch that the living room audio market is much smaller than it once used to be, and that music is now a very personal experience that tends to be consumed through headphones more than through speakers, but I may be wrong. If this is a market worth pursuing, then smart speakers could be marketed as a replacement product, which I think is Apple’s initial strategy.
In Geoffrey Moore’s “Crossing the Chasm”, the author advocates a whole product strategy. To do this with smart speakers, you need to own the speakers and the final product which the customer will pay for. Amazon + e-commerce is one example of a whole product, except that it just isn’t very compelling – the smart speaker just becomes a glorified Amazon Dash button. My point in the post above was that Tencent (WeChat) has more breadth in services that people actually pay for, so they might be able to find some good solutions. Furthermore, if you recall the recent “bot” hype, that’s actually a better fit for a voice UI, because that was supposed to be conversational as well, and it was also centred around messaging apps.
LikeLike