Top news stories for Episode 11 (September 14, 2017):
1) Red Alert: Amazon looks to disarm critics of its $100M "Alexa Fund" in aftermath of Echo Show-related conflict; Ring CEO compares Amazon involvement to "nuclear radiation"
2) #VoiceFirst in China: Amazon looks to take on Alibaba by staffing up.
3) Google and LG team up to give smart appliance buyers a free Google Home through October 9, increasing their market penetration.
4) More Microsoft convergence, as MS teams up with Facebook to develop a common framework for machine learning.
5) Voicebot.ai Story Of The Week: 62% of the over-10,000 Alexa skills have zero user reviews, while four of them have over 1,000. Is this a roadsign pointing the way to Amazon long-term dominance, or a sign of hidden weakness in the Echo-system?
This Week In Voice available via:
YouTube (+ closed captioning)
Panel for Episode 11 (September 14, 2017):
Leor Grebler is CEO and co-founder of of Unified Computer Intelligence Corporation (UCIC), a company dedicated to bringing voice interaction to hardware. UCIC's goal is to make this voice interaction with technology more human and natural.
Bradley Metrock: [00:00:10] Hi! And welcome back to This Week In Voice for Thursday, September 14th, 2017.
Bradley Metrock: [00:00:18] My name is Bradley Metrock, and I'm CEO of a company called Score Publishing based in Nashville, Tennessee. Our sponsor for this podcast is VoiceXP - blazing the trail in voice technology. VoiceXP is taking the lead in developing Alexa skills for some of the best brands in the world. With VoiceXP, all you have to do is say it to revolutionize your marketing strategy. And I fully encourage you to check them out at .
Bradley Metrock: [00:00:48] We are thrilled today to be joined by two excellent, excellent guests. First is Leor Grebler - Leor, say hello.
Leor Grebler: [00:00:56] Hello, Bradley, it's great to be here!
Bradley Metrock: [00:00:58] Leor, thank you for joining us today. Leor is CEO and Co-Founder of Unified Computer Intelligence Corporation - UCIC for short - a company dedicated to bringing voice interaction to hardware. UCIC's goal is to make this voice interaction with technology more human and natural. Leor, thank you very, very much for sharing your time and insight with us today.
Leor Grebler: [00:01:20] Thank you!
Bradley Metrock: [00:01:21] Our second guest is Matt Hartman - Matt, say hello.
Matt Hartman: [00:01:24] Hi, thanks so much for having me!
Bradley Metrock: [00:01:25] Thank you very much for joining us. Matt Hartman is a partner at Betaworks Ventures, and also runs the really cool weekly voice technology newsletter, Hearing Voices. Both those Web sites are linked off of our page - Hearing Voices, I believe it's HearingVoices.XYZ - it's a great newsletter, you should definitely check that out. Matt, thank you very much for joining us and sharing your time and insight with us today.
Matt Hartman: [00:01:49] Thanks again. I'm excited to talk voice.
Bradley Metrock: [00:01:52] Yeah, absolutely! We've got a lot of different things going on. Also, right off the top - and it will be our last story of the week - but we are pleased to announce that we have a partnership with VoiceBot.ai, which is another really great resource for folks interested in voice. I encourage you to check them out. We're going to have a VoiceBot.ai Story Of The Week, each week moving forward, which will take one of their stories that they have reported and highlight that for discussion. So we wanted to make sure to announce that as well.
Bradley Metrock: [00:02:24] With that - let's get to the news. Our first story this week is one that's sort of been growing and festering a little bit over the last few weeks, and it came to a head with the big CNBC article which we've linked off of the news page on ThisWeekInVoice.com. Amazon is looking to - "disarm" is the word on the page - muzzle, reason with - critics of its $100 million Alexa fund in the aftermath of a conflict they had with the company that had some technology similar to the Echo Show. In this article, it's very interesting, and I think rather unflattering: the CEO of Ring compared Amazon's involvement with the Alexa fund to nuclear radiation < chuckle > in that - you need nuclear power, and it's a productive thing for the world - but you don't want to get too close because you don't know what's going to happen to you, was the point he was making.
Bradley Metrock: [00:03:24] Matt, I'm going to start with you on this. How do you look at the Alexa Fund yourself, personally - especially with your role with Betaworks Ventures? Share with us your thoughts as you look at this article.
Matt Hartman: [00:03:38] We do a lot of work with Amazon and Alexa and know the Alexa Fund. I think there's sort of - there are a few different ways to look at this. On the one hand I sort of find it unrealistic that Amazon got a bunch of proprietary insights from making an investment in a company that looks like this. Did they kind of do something that was not right?
Matt Hartman: [00:04:03] And the second is - just generally, does it make sense to work with a corporate investor? Many corporations have venture capital funds. They're investing often because of a particular strategy. In this case it's most likely, my guess is, that Amazon said we want to make sure that as many voice-activated products use Alexa as possible, and I think they're pretty transparent about that. I am sure, on the other hand, that as an entrepreneur going in - you don't know how people in the company are going to want to interact with you. And so I think asking those questions up front and just having a transparent conversation about that probably helps a lot.
Matt Hartman: [00:04:40] There are some disadvantages to having a corporate investor; there are strategy changes over time. This year it may look like one thing, and next year they just may have a corporate decision that says we're not gonna do this any more, it's not relevant to us any more. One of the thoughts I had as I was reading the article was that if you believe that Amazon - I hadn't thought about this before - but if you believe that Amazon could be a competitor - that is a risk in starting the company, period. A question to ask yourself - it's kind of interesting: all things being equal, would you rather have them potentially be on your side or not? I can make the argument either way. I think it's kind of up to the founder, and up to how much money they are able to raise, and whether they think there is so much platform risk (which is, I guess, another way to think about this) that it's worth having the corporation where the risk is associated be an investor.
Leor Grebler: [00:05:29] That's actually a really good point, because we faced that exact same issue a couple of years ago when we were working on the Ubi, which was a product I guess similar to the Amazon Echo. And then Amazon came out with the Echo, and we had to figure out whether we were going to try to compete or whether we were going to try to potentially work with Amazon.
Matt Hartman: [00:05:52] And in this particular case I'm not sure that the technology was the issue. I think that the key advantage Amazon has is - they have this retail site where they can put the product that they made on the front page and sell a ton of them. Right? So they put their products ahead of everybody else's. I mean, they've been fighting with Google over Chromecast, and that's actually a big piece of where the advantage comes, having nothing to do with the kind of technology.
Bradley Metrock: [00:06:14] Yeah. And Leor, thank you for reminding me about your history with Amazon as well, as far as the Ubi and the Echo is concerned - and anyone who wants to hear more about, that check out that episode of The VoiceFirst Roundtable.
Bradley Metrock: [00:06:26] I think y'all's answers really capture sort of the murkiness of this. Matt, as you said, I think you're right - there is no right or wrong answer here. It's just very specific to every company's needs whether they want to take money from the Alexa Fund, or seek it, or whatever. I do have a question that maybe one of y'all would know the answer to: if you're going to get investment from the Alexa Fund - are you able to do that if part of your strategy is having your cross-platform voice application?
Leor Grebler: [00:06:55] I don't know from any inside knowledge at all whether that's the case, but we could look at a specific example with Kitt.ai - they were an investment from the Alexa Fund and they exited to Baidu - so that definitely didn't stop them from being acquired by what could potentially be a competitor to Amazon and the voice-base.
Bradley Metrock: [00:07:18] Interesting.
Leor Grebler: [00:07:19] And at the same time, though, they're still functional - they're still providing voice technology that could work as the Alexa trigger word. They provided trigger software, but their stuff could also work completely independent of Amazon, so that definitely didn't affect them at all. The other thing, too - I think we tend to treat Amazon like it's one big conscious giant. When the reality is that there are a lot of players within Amazon, a lot of competing divisions, and I would see the Alexa Fund as being fairly independent from the rest of Alexa - even though it's sharing the name Alexa. So I think it is still possible to get involved with them and work on a strategy that is also independent of Amazon, but potentially bring a lot of cachet to the companies that are working with the fund.
Bradley Metrock: [00:08:16] So all in all, at the end of the day, Amazon having this Alexa Fund is a positive. Would you all agree with that?
Matt Hartman: [00:08:23] Right now, having more investment in the category will create more potential products, and that's good. I do think that the story is probably not a positive for founders who are sort of debating whether they want to to raise money from them, but it could be a short-term thing also.
Leor Grebler: [00:08:41] I think it's definitely a good thing - more so than any potential negative drawbacks. It's great to have more money flowing into voice to develop new products.
Bradley Metrock: [00:08:51] Cool. OK. So we'll leave it there, and yeah, that's my sense as well. Amazon - as we've spoken about before on this show - for Amazon to be leading voice technology in the way that they are, because they do have a unique level of trust with the consumer that some of the other companies don't have - cultivated a lot of trust with how they interact with customers and all the things they've done innovating for customers. And hopefully they don't do anything with its Alexa Fund that compromises that reputation.
Bradley Metrock: [00:09:21] We will move on to Story #2, another Amazon story (imagine that!) - Amazon is ramping up in China to take on Alibaba. The article that's linked here talks about a number of personnel types that they are hiring, including engineers for Alexa. Leor, I'll start with you on this. What did you take away from this article? Do you think Amazon will be successful in taking on Alibaba? How do you see this playing out in the overall war for voice?
Leor Grebler: [00:09:53] I had kind of a mixed reaction to this. My initial feeling was that this could just be more of a defensive strategy to keep Alibaba on its own turf and having to worry about keeping its market share there and not coming to North America or some of the other geographies where Alexa is being released. Unless Amazon were to go out and acquire a partner in Asia, I think it will have a lot of difficulty getting into those markets. I mean there are a lot of competitors that are already there in voice, and have been doing it even before the Echo was widely released. It seems to me it's more of a kind of defensive strategy to keep them out of North America.
Matt Hartman: [00:10:49] That's interesting, I hadn't thought about that - about the defensive strategy piece of it. I sort of just assumed there is massive growth happening in China. Amazon is a company that's never shied away from trying to grow. They've done a really amazing job. And so I just wasn't surprised about that. I did wonder, on the voice side - if you believe that to create a great voice assistant, a great voice interface, you have to have a lot of data and then learn from it over time and adjust your algorithms - I feel like the language issue - I wonder if they're starting from scratch? I'm kind of kind of curious if you guys had thoughts on that?
Leor Grebler: [00:11:21] Oh, yeah, they would be far behind in terms of being able to collect samples. When Amazon was developing Echo, they had a few acquisitions that kind of helped move them forward really quickly in getting technology that would help for English-based voice interaction. What I've learned is that they've done a lot of work in capturing of voice samples. They had a large project that involved using open source or openly available tools. But having to go and do the kind of the grunt work of collecting samples - if you look at Biodata, look at Alibaba - I mean, they've just been in the space for so long - look at WeChat - just using so many more voice samples, probably billions of additional ones every day. And with all the different dialects, I think it's going to be a bit of an uphill battle. Because you have to get to that level of word error rate for the voice interaction to be reliable enough for people to just put it in their homes and forget about the ? product grabber blend in the background. So you're right, they have a lot of work ahead of them.
Bradley Metrock: [00:12:27] Moving on to Story #3: Google and LG are teaming up. I believe this has started now and it's going through the first week or so of October. What they're doing is for everyone who buys an LG smart appliance, Google is giving that buyer a free Google Home. And this is just another one of these marketing strategies designed to increase the penetration of the Google Home device. We're seeing a bunch of this. And Matt, I'll start with you. Is this a good thing for Google to be doing? Is this a bad thing? Are you in favor of strategies like this? We're seeing a lot of them. What are your thoughts?
Matt Hartman: [00:13:08] To me this makes sense in the context of Google. Google's business model has put a search box as close to your face and now as close to your voice as possible. I believe they're the number two player. Alexa is clearly ahead. And so if they can get more devices out, more people using it - more searches will happen that are powered by Google. And I think it's a smart move. I think it would be equally a smart move if they figured out a way to get the Google Assistant into a car, because when you're in a car you might want to do searches too.
Leor Grebler: [00:13:35] This is just a marketing gimmick. Get something for free when you buy an LG product. Trying to get them out as much as possible. But LG has just been everywhere when it comes to voice. So I don't think it defines a true partnership between LG and Google in any way just by them marketing this, because they've looked at putting Alexa into products. I remember, it was last year at IFA, they had a product called SmartThinQ, and this was supposed to be a home hub. It was originally going to have its own voice interaction component to it. But in the end they just said "Nah, we'll just put Alexa on it" and they made a big announcement that they were going to have Alexa voice service integration in the product; they continue to make announcements around that. But I don't think it's actually ever been integrated into a product. What would have really been kind of a big thing is if they came around and said "all LG appliances are going to have the Google embedded Assistant in it" - it's not you get a Google Home product and you can then control your appliance. It's like it's built into the product. That would have been much more of a signal that OK, they're finally serious about partnering with the company on voice.
Bradley Metrock: [00:15:00] I want to follow that up by asking a question that we've discussed on the show several times: in a year or two, what does the average, the median home in the United States look like in terms of these smart speakers? We've had different people on the show and there's all sorts of different opinions on it. But it's my sense that especially in the aftermath of Amazon rolling out the functionality where you have multi-room audio (this is just obviously recently, a couple of weeks ago) - you can link as a group all of your Echo devices to have multi-room audio, which is really, really a phenomenal feature. So the question becomes: why would you buy a Google Home unless you're invested in the Google ecosystem? You're not just going to diversify and have different ones, because why would you do that when if you just need different ones for each room, you're going to want them all to be part of the same ecosystem, are you not? Because then you can take advantage of multi-room audio or perhaps other features that come out down the line. So my question is: in 2018 going into 2019 - are most houses going to have a Echo Dot III and a Google Home II and a Harman Kardon Cortana-powered speaker along with the HomePod? Or is Alexa going to continue to move and end up winning this thing? What's y'all's vision for the future of what smart speakers are going to be in our house?
Matt Hartman: [00:16:28] If you look at historically how technology sort of rolled out its start - the best experiences tend to start out fully integrated. The experience of having two Google Homes in your house OR two Alexas in your house is better than having one Google Home and one Alexa. Because those don't talk to each other as easily. So I think that in the near term, the best sort of fully integrated experiences are going to come from a single brand. But that doesn't necessarily mean that Google wins that or Alexa wins that. But I think when you walk into a home, they will have a consistent company that they are using. I think it's worthwhile to think about the company's overall strategies - because who's going to make the assistant that is best for consumers? If you look at Amazon's business model, they sell you things - which is, they make a really good experience for selling you things. Google's business model is answering your search questions and creating a really good consumer experience, and then charging people on the other side of that. And so a question is: which one of them is going to be better at the machine learning that gets really good at recognizing what you want? Which one of them is going to have the incentive, or disincentive, to make a skill or an action that gives you what you want and doesn't necessarily sell you their own product?
Matt Hartman: [00:17:44] I would have a hard time picturing Amazon being ok with the Wal-Mart skill being on Amazon - whereas Google is indifferent; you can either buy from Amazon, you can buy from Wal-Mart, you can buy from I-don't-care. If you look at what Amazon did with Amazon Web Services - they were very democratic about it. It was a different - they didn't care if you were building an e-commerce site on Amazon Web Services; they let you do it. And so the question is to me: Are they going to have to take that much of a sort of independent approach with Alexa? Because if they don't, I think there's a lot of incentive for them to block things that are competitive.
Leor Grebler: [00:18:18] That's definitely a good point, especially with households picking one product over the other. Just from a user experience, it's annoying to have to refer to multiple personalities in the home - whether you're going to say "Alexa!" - no no, this one's a Google Home - and then you have some other product - "Hey Cortana!" or "Hey Siri!" for HomePod. I think you're absolutely right, Matt. When a household picks a product, they're going to probably just expand on that given product.
Leor Grebler: [00:18:50] From a device side, looking two years down, I think we'll continue to see Amazon and Google and Apple manufacture products for voice interaction. But we're going to see them more and more displaced by third-party hardware that is integrated with either embedded Google Assistant or Alexa voice service. And what could be interesting is that we end up having products that can have those multiple personalities that can be set to one or the other, or potentially both at the same time, marketed to to consumers - in which case, the war might actually be fought for the different hardware makers to put in the service for Google or Amazon into their products. So maybe there'll be some incentive for a company marketing one integration over the other. That's where I think there is going to be an interesting war zone.
Bradley Metrock: [00:19:56] Moving on to Story #4. A couple of weeks ago we reported on the big news that Microsoft and Amazon had partnered together to allow Cortana and Alexa to talk to one another, and all the different doors that that framework opens up for development and for voice interaction. Well, we've got another story along those lines this week - with Microsoft teaming up with Facebook to develop a common framework for deep learning, machine learning, whatever term you want to use. And my question for the panel - and Leor, I'll start with you. It's two parts. Number one, what did you take away from this article? Is this a good thing, is this going to be helpful, or is it sort of just noise? The second thing is - I've commented sort of negatively on Facebook on this show before, primarily because I don't feel like anybody trusts that organization very much. Is that identity of Facebook going to get in the way? Or are Microsoft and Facebook doing really important work here?
Leor Grebler: [00:21:03] Yes, I'm really confused by Microsoft. I don't know - I'm trying to figure out a larger strategy for them and what they're up to because they've worked on their own stuff and they've worked with others. They wanted to have this Cortana speaker with Harman; Harman was bought by Samsung. They also acquired Maluuba for natural language understanding. They have this Bing speech-to-text that they put up for their own services - it's kind of like there's a lot of shots that are going out - and again, it was Microsoft and Alexa and Cortana speaking together. The conclusion was that, hey maybe this could work for a lot of office applications. But, this?? I don't know - I think maybe this has to do with Facebook doesn't have a lot of services that are designed for enterprises, like an AWS or like Google App Engine. Maybe this is their attempt to try to expose some of their their infrastructure as a service to take away potential areas of learning from Google or Amazon.
Matt Hartman: [00:22:26] I'm glad Leor got this one first, because I always look for sort of the competitive strategy reason why people were doing things - and I had a "whose interest is it?" for something to work. It's obvious when Google does Android that what they're trying to do is compete on a different layer - kind of commoditize what Apple's doing with iOS. I had a hard time seeing really what either company gets from this. If it's open standard AI, I guess you could argue that Facebook has data - so there being more models is good, because they're the ones with the data and they can use that. But I don't know - I had a hard time seeing this one also. I do think that Microsoft seems to be putting together an interesting strategy around providing the kind of the toolkit layer of what people are going to need in this conversational economy. And I mean conversational both in text and voice, and a lot of developers like the software they're building. I feel like they have a cohesive strategy on that side. I just don't totally understand why that would be consistent with promoting, or with joining to promote, kind of an open standard - other than if it benefits that kind of core strategy in some way. And it's just also noticeable who ISN'T in that list. If it's Microsoft and Facebook, it ISN'T Amazon and Google. And so, why? They just decided the two of them made sense? Or is this a competitive thing that kind of helps them deal with with some of their competitors?
Bradley Metrock: [00:23:58] Story #5, which is our inaugural VoiceBot.ai Story Of The Week. This is a story that Bret Kinsella over at VoiceBot.ai surfaced, and it's very interesting. They took a look at all the Alexa skills that are available now - I think the number is like 12,500, something like that. Sixty-two percent of them have zero user reviews. None. Not any. Well, four of them have over 1000 user reviews. And at least the initial question for this discussion is: just give me a one word, yes or no, as to whether this finding is a positive for Amazon. And then elaborate on how you think this is going to play out, and what you think this means in the longer term. And Matt, I'll start with you on this one.
Matt Hartman: [00:24:52] So my one word is: irrelevant. I don't think it matters. We built a product at Betaworks a while back called Homescreen, where we had people screenshot with apps from Homesceen, it was an IoS. So the question was what new apps can we see that people are using and will they share with each other. What we found was that we tracked 50,000 unique apps; of the 50,000, nine hundred of them were on more than one person's Homescreen. So what that means is that basically there's a handful of apps everybody uses, and then everything else maybe you use one here or there. But if you think about Amazon's strategy - what they want to make sure is that you use Alexa enough so that when it's time to buy something, you use it to buy an Amazon Services unit. Everything else is just top of the funnel. And I think it's very comparable to Apple's business model being: we sell you phones, and we're going to put a bunch of apps on there that we think are good. If somebody else builds a better one, great. But that's like premium for us, in the same way that Apple benefits from there being 50 good apps that I care about - whether that's Twitter and Instagram and Facebook, that everybody cares about - and then the longtail of millions that maybe a few people like. And hopefully there could be an economy there. That kind of is analogous to me of the reviews in Alexa. I happen to use Alexa mostly for - I think many people use it mostly for - its default features right now. I do think there will be some killer apps in there. But I don't think it is a negative signal that some of the apps - many of the apps - or even most of the apps - aren't killer apps.
Leor Grebler: [00:26:20] My one word answer - it's negative. It's not good that so many apps, so many skills that have been developed are not getting any users at all accessing them. But it kind of reflects a larger problem, which is around skill discovery, and even accessing skills. This is a problem that Amazon will need to address and fix some time over the next year. I'm not surprised that you have ... from Matt's example, there should be a fatter head for skills versus apps, because just discovering a skill takes a lot of effort. You have to go to Alexa skills and it's not the way that you're normally accessing the device. You're not asking "Alexa, what are some great skills that I should access?" Typically it's because you found out about it, or it was just that you needed something and it was such a problem that you had to go out and you had to research what was the potential way of doing it.
Leor Grebler: [00:27:33] The other problem is just the whole interaction with skills. Right now, you have to invoke the skill in order to access it. And this is different than, I guess, the approach that Google took with actions - in that if someone subscribes to an action, they don't necessarily need to invoke it in order to access it or subscribe to a skill. And this is really the kind of approach that maybe Amazon should take: "Hey, developers - go out there and develop something!" But we're going to try to drive people to access this particular skill you've developed because it's something that could be really, really useful. I don't think that there's going to be a million skills for the Echo. I think there's a natural limit for skills that we're going to reach over the next few years, at least as they are today. Amazon has been doing a lot of effort to build a developer community and to really promote things and to run contests and to make skills more interesting. And I think this improves the overall quality of the skills and makes for a longer tail. There's a lot of issues that I think this particular article starts to shed light on.
Matt Hartman: [00:28:53] I agree with Leor - I agree that...it's kind of a question for whom is is it good or bad. So for Amazon in the short to medium term, I don't think it's bad. I think that people are using it for Spotify and are using it for cooking timers and using it for the weather and the news.
Leor Grebler: [00:29:14] That's true.
Matt Hartman: [00:29:15] But I do think - I agree with you totally - that for people who are developing for voice - the entire infrastructure around having intents, and basically just getting a binary ... well, not binary ... but a response like: OK, here's the intent that was triggered, or nothing was triggered ... versus actually getting to learn from how our users are behaving - is a negative for voice developers because they can't make as good experiences.
Matt Hartman: [00:29:38] And so I think it kind of goes back to being bad for Amazon - the longer term when we go back to thinking, OK, five years from now what are our houses going to be outfitted with? Probably not with the company that has not made it easy to create the most useful things. I kind of think about it as not mattering to Amazon's core business right now. It may matter to whether they ultimately win. But I think it's almost more of a reflection than a driver.
Bradley Metrock: [00:30:06] This is an interesting one! And I think the voicebot.ai article does a good job of showing the shades of gray on this. From one standpoint, if you're going to compare what's going on with Alexa skills to the Apple app ecosystem - I think that's a pretty apt comparison, and really one of the main comparisons we have to make. It's not like there've been a bunch of companies in the history of the planet that have done something like this. Apple's obviously been the number one example. But the problem becomes - the example cuts off halfway, because you have no monetization. So you've got all these skills, and you've got this Apple-like environment, or it's trying to be an Apple-like environment in many ways. But wait a minute: sorry, you can't sell your skill. And so I think that impacts the market to a huge degree - and it will vary based on a different genre or a different type of skill you're talking about - but an overall net effect will be a reduction of quality.
Matt Hartman: [00:31:07] That's an interesting point. Can I make one kind of observation about it?
Bradley Metrock: [00:31:10] Go ahead.
Matt Hartman: [00:31:11] Apple is, in a way, an anomaly in the discovery side...so I totally agree about the discovery being an issue. In the iPhone, you find apps for it by using the iPhone. In Windows, you found apps for it by going to, like, CompUSA and buying software off the shelf, or like AOL mailing you a CD. In Alexa, I would say that finding a skill is much closer to the CompUSA example. Right? You have to go into the Alexa app, which doesn't really make sense. And you can enable it on the device but you can't really discover. It's hard to describe a skill using words ... you've got to listen to it .. and discovery is sort of hard on voice. We saw the same thing happen to chatbots - where people sort of assumed that if I was featured in the slack store, or if I was featured in the Facebook Messenger app, that means that a lot of people are going to go there to find the thing. And the reality is that - it's much harder. It's much less likely that they're going to find the thing if it's not natively in the product that they're using. And so, what I've seen at least, is that the companies that seem to be doing really well own their own marketing. If you own your own marketing, you're less reliant on sort of this theoretical or hypothetical organic discovery directly from being, like, featured. I agree with you that the analogy on the iPhone is flawed in that way.
Bradley Metrock: [00:32:32] It doesn't help anything that Amazon has not communicated its long-term monetization strategy. Right now all we've got is "Hey, you - you're making an Alexa skill for productivity." Or like the Earplay folks: "You're making an interactive experience." Or a game. If you do well enough, guess what? A check's going to appear in your mailbox. If you don't, a check's not going to appear in your mailbox. That's pretty much where it is right now. And it's replication in many ways of the Kindle publishing ecosystem that they put together, which in similar ways only serves a very small amount of people and frustrates many others. So it'll be interesting to see if they evolve from that. But certainly they're not communicating on this topic at all. And that hurts.
Bradley Metrock: [00:33:25] The other thing I want to mention here, too - it's important to note: 62 percent have zero user reviews. That doesn't mean zero people enabling it. And actually, the voicebot.ai article, I believe, mentions that there are some pretty well-known Alexa skills that have a very small amount of reviews related to how many you would think they would have. And honestly, if you asked me how to leave a review for an Alexa skill - I wouldn't have any idea how to do that. I guess you do that in the back end of it. So all of that is to say - I think most people would agree if you have zero user reviews, you're not getting a lot of action. But still - we don't know everything that's going on there - but from where I sit, I think, just to put a button on this: I like all the different skills. I hope there are a million of them. I'm an entrepreneur at heart and I like to see this entrepreneurial aspect of it. Because if you're going to have a million skills, over half of those are going to be individual entrepreneurs trying to dive into the space and create something interesting. And I love that. I hope that that continues. But absolutely, in terms of discoverability and some of this other stuff - they've got a lot of work to do.
Bradley Metrock: [00:34:44] So - gentlemen, thank you very much for joining me today. This was fantastic. Thank you for sharing your time and your insight. You both are two amazing guests.
Matt Hartman: [00:34:54] Thank you so much for having us.
Leor Grebler: [00:34:56] Yeah, it was a lot of fun!
Bradley Metrock: [00:34:58] So, for This Week In Voice, September 14th, 2017 - thank you for listening. And until next time.