4mation presents a panel discussion on why voice technology matters in 2018.
Voice assistants and voice search have arrived and they’re here to stay. Join us at 4mation to hear insights on voice technology from a panel of experts. They will discuss why voice matters, and why it should be part of the next stage of your digital strategy
Jason: Good morning everyone, and welcome to our State of Voice 2018 Event. My name is Jason Hincks and I’m the CEO here at 4mation Technologies. For those of you who are new to 4mation, I’ll just give you a little bit of information about our organisation. We’re a full-service digital agency with a commitment to the development of local talent. We have 60 of the country’s best developers working in an agile environment. We’ve been around for 17 years, which is a long time in the tech space and through that time, we’ve gained plenty of insight into how businesses can leverage digital technology to drive business outcomes, which brings us to today’s topic. So, like most of the conversations we have here at 4mation, it started with our clients. Our clients were asking us if they should be using voice as part of their digital roadmaps. So we did some research and as we scratched the surface, we began to understand just how important the technology of voice is going to be. In fact, ComScore recently published data suggesting that by 2020, 50% of all searches will happen by voice.
Like most in the room, we’re fairly early in our journey, but I’d like to share with you a little bit of the insight that we’ve uncovered so far, and it focuses on two areas, the first one is utility. Utility centres on creating rich user experiences through voice. While availability, focuses on leveraging voice to help your customers find you and having meaningful conversations with them when they do. We’ve captured these insights in more detail in a report which you have on your seats and we’ll also make that available as an eBook post event, but you haven’t come here to listen to me speak. You’ve come here to listen to the experts.
So, without any further ado, I’d like to introduce those experts. First of all, I’d like to introduce Nima Vadiee from Amazon. Nima is the Head of Solutions Architecture, Asia Pacific at Amazon Alexa skills team. A tenured Amazonian, Nima has helped brands across Australia and New Zealand engage more customers through voice with the Alexa skill service, and has been instrumental in the launch of Amazon Echo across seven different countries. Nima has spent most of his career working as a software engineer and developer advocate and he’s passionate about enabling developers to build compelling voice experiences. Welcome Nima!
You will have noticed we had a little bit of a change in the line up today. So, originally, we had Mark Thomas who was speaking, but, when we were made aware that Nima was available, we jumped at the opportunity to have him here and, I’m sure he’ll give us some great insight. I’d next like to introduce 4mation’s Head of Delivery, Scott Morris. Scott is one of the founders here at 4mation. Scott has overseen countless software projects for clients, including ABC, SG Fleet, Grace and Kennards Self-storage. Welcome Scott!
And last but not least, we have 4mations Resident Digital Marketing Strategist, Regan McGregor.
I’m going to make myself comfortable on these quite uncomfortable stools. Is it possible to be comfortable on a stool? So as our guest, Nima, I thought I would direct our first question at you, why is voice becoming such an important tool for business?
Nima: I actually did a little bit of studying of your pamphlet right before this, so I’m prepared for this question.
Jason: These aren’t prepared questions.
Nima: Actually, when we first started this in the U.S., it was a much harder conversation to have with businesses where it was a conversation around like, you know, this is maybe an innovation play for your company, just dip a toe in and build something that can be a proof of concept that you can sell internally and, maybe we were just you know, we wouldn’t say like this, but maybe they would see some value out of it, maybe they wouldn’t. And it was – most brands are pretty timid, but the reception was warm because; building and deploying something was fairly simple lightweight, they quickly you know there’s a lot of experts in the space that were able to leverage other existing skill sets like mobile developers to help you deploy these voice experiences, but it’s really evolved quite quickly. And I’ve been lucky enough to see that evolution from the US, UK and now in Australia.
It’s really become the next medium, the next place where companies will interact with customers and/or are interacting with customers. So, the statistics that you see, like it’s happening, like these brands are actually seeing success and engaging customers in an ongoing basis, making it part of their daily routines, sticky, simple short interactions that are natural and really comfortable to users. So, the successes, I think it’s here, it’s just a matter of kind of embracing it in each new market.
Jason: Yeah, it certainly feels like voice is here, maybe a question for you Regan. So, for organisations that are looking to start their journey, where should they start?
Regan: Yes, I think one of the easiest places to start to kind of start building presence is in voice search, especially around local search. So, there were some statistics out of Google that every year-on-year 130% increase they’re seeing in searches such as personalised ‘around/near me’, so location-based, and I think that provides a real opportunity for businesses to be seen where they are. So, by optimising your – My Business listing or your Bing business listing or providing information like your trading hours and the payments you accept, is a really easy way to get started and start optimising for voice.
Jason: What feels like the place to start, is in a quite functional space, which is not surprising. Scott, where do you think the opportunities are for brands?
Scott: For brands, I think as a first port of call is to get your voice search optimised, you need to be ready for that because voice search is here, it’s already very popular because it’s so easy to use and it feels natural. But beyond that, I guess similar to what Nima was saying, here in Australia, it’s still fairly new. So, for brands, I think it’s a good opportunity to position yourselves as innovators, as a brand who can keep up with consumer technology and a brand that is in the same space where your digital native customers are, I think that’s really important, and we’ve already seen brands in Australia do that. You know, you’ve got Qantas, you’ve got Uber, you’ve got some of the banks, you know, they’ve dipped their toe in the water. As Nima said, a lot of them are still fairly small and you know, ‘proof of concepts’, but I think just being there right now is a big opportunity to be a leader.
Jason: Yeah! It just feels like I’ve talked a little bit about it there in my opening, about finding sort of having voice be the mechanism to find you, but then having a meaningful conversation with your customers, can you talk a little bit about the customer value in that equation.
Scott: Yeah! So, there are two main things that I see, where voice really does a good job. So one is convenience, just that hands free, eyes free, frictionless ability to interact with that voice interface, so if it’s a really quick thing like ordering your favourite pizza, you know, via Domino’s, and that’s possible right now, whether it’s booking a table through Dimmi, again, that’s doable right now. It’s those quick little frictionless opportunities to interact with the brand, where voice can work really well. Another area that I see that can work, is a new kind of content marketing. So, an example I’ve seen from the U.S. is with Tide and they do kind of carpet cleaning solutions and they built an Alexa skill to answer the question, “how can I get this stain out of my carpet?” And you know, that had answers to over 200 different types of stains and I think that’s a good example and a neat example of providing useful content for the user while still pushing your brand.
Nima: Simple but effective like that, that one took all of us by surprise by how popular it became. It’s just something to people in their day to day lives or question you would ask like, ‘how do I get a grass stain out?’ And it’s really, really useful on an ongoing basis.
Jason: I think the question that the entire room wants to know is, is there an answer to the question, should I eat this Tide pod?
Nima: I hope there is an answer to that. I haven’t tried
Jason: It seems to be a scary trend. So, Nima, you’ve talked a little bit about the fact that lots of brands are embracing voice and you gave a couple of examples of organisations dipping their toe in the water, but who’s doing it well? like, you know, even though we’re in the formative stages, who would you point to as being the sort of shining lights in the process so far?
Nima: Yeah! I mean I should clarify that statement. I meant they were dipping a toe in the water in the very beginning. We’ve definitely moved to the point where it’s fully embraced by a lot of organisations. And we saw this actually at CES in 2015 versus 2018 where you’d meet with one developer from an innovation group, as opposed to the next time you may view this company three years into their skill being live, the voice experience being live it’s, you know, you have 20 people around in the room ideating on different ways to build something really successful. The ones that have done it really well, or again, the skills and the voice experiences that are simple, easy to use and sticky on an ongoing basis, you know, one that may be our most notable one, that is always popular in every market is jeopardy, which is just a game show in the U.S.
And then from there, but the ones that just provide utility calling Uber or ordering in your favourite meal, reordering something, ordering a pizza from Domino’s, reserving a table. And then just some of the more silly or fun ones; a lot of the trivia ones, jokes, simple things that can actually help you, like remembering what to order from a supermarket. For example, in the UK, there’s a supermarket chain called Ocado, that does online delivery, people just want to become quickly become one of the experiences that the most popular, just adding and reminding yourself to pick up something or throwing something into your cart to order later. And then the one-use case that is still the most popular is streaming audio. So there’s a massive and amazing piece of technology that’s still being used as a radio.
Jason: Yeah, it works. So I think that’s really exciting that we are talking about, you know, the experience and the journey that brands are taking, kind of being anchored in what the sort of personal requirements and the personal journey is, I’d be interested to know from the panel, just, you know, maybe an example or two from each of you, how you use voice day to day. Maybe we start with you Scott.
Scott: Yeah! So, you know, I use it a lot. I use it for all the typical stuff playing music. Managing shopping lists, so adding items to the shopping list, weather, reminders, timers, you know, that’s the stuff everyone uses it for. I’ve also used it to check if a business is open or closed, you know, before I head out. I use it to see when I need to leave by a certain time to get to a particular destination. I’ve also programmed it so that whenever my three-year-old son asks ‘when is it bedtime?’ the time is now, the answer is always now.
But again, I think what I find really great is just being playful with it, so just trying things out.
So, as an example, I was out at Westfield, two kids in tow and I’d finished my shopping and I realised I didn’t remember where I parked my car, so I whipped out my phone and gave it a shot, where, you know, I just asked it ‘where did I park?’ And sure enough, it just came back. I think you’re parked here, came up with a map and there it was. So, I think just having that ability to be playful with it and discover those new functions will help people with this service discovery, which can be a bit of a challenge, but you know, I imagine that’s all really important, I imagine that the platforms are kind of aggregating all of that information and knowing what customers want and guiding that to what to build next.
Jason: I say it sounds like the bedtime function alone has your voice assistant paid for itself already. Nima, how about you?
Nima: So, I live on Bronte Road and there’s about 58 different buses that can take me to Bondi Junction, it seems like I’ve finally gotten the hang of it, but in the morning I just check, I use the Next Transit skill to tell me which bus I should be getting on, what time it leaves and it gives me just enough time to get there, you know, right as it’s arriving. And then a whole different story, I’m flying to the U.S. next week and I was checking Qantas, you know, checking the flight status, gate information, those types of things. It integrates really well with just my day to day activities. In the morning, listening to the news and listening to music while we cook dinner. I honestly do use it for cooking quite often or just getting recipe ideas. I’m probably an example of someone that’s embraced it fully, given the amount of time I’ve been used to it, but there’s just a lot of different options.
Regan: I’m a big user of the technology and usually I find that it is most helpful when you’re in the middle of something, driving or in your living room, having a conversation and you just remember something and you need a quick way to get it down. And I can think of the most notable example at the moment. Me and my fiancée, we are planning a wedding which I highly advise against [inaudible 15:04] [laughter] and I was getting into lots of trouble because I kept forgetting stuff. Like at the beginning I needed to call this person, call this person, I need to do this. So, I found this handy app called Todoist and it has an app that goes across my phone, it’s on my watch, it’s on my computer and across all the voice devices. So, whenever I’m driving the car or I’m at home sitting on the couch, I can say add this to my to do list or, ‘what’s due today?’ and it will give me the information I need and so I’m getting in a lot less trouble.
Jason: Again, paying for itself, I would say at this point. I talked a little bit about in my opening about the requests that we had from our clients, Regan, and a lot of those early conversations were around process automation and it’s probably not the thing that we intuitively think of when we think of voice. Talk a little bit about the opportunities of process automation for voice.
Regan: Yeah, so, I think there are definitely opportunities for businesses to use this. Non-consumer facing and one notable thing I can think of, coming from an e-commerce background, we can’t all have robots in our warehouse like Amazon. So I thought, what if you could use Alexa to communicate when your hands are busy, when you can’t touch a screen, so I imagine like a forklift driver could be driving to the pallet that they need to be picking up, and they just ask Alexa, ‘where is my next door located?’ they go down, and grab it. Or even somebody on a picking bench could ask Alexa what’s in this next order, and then make it off as shipped and it just provides some utility there in the fact that they don’t have to bend and turn around and possibly even injure themselves from manual kind of actions and just makes life a heap easier for them.
Jason: So, it sounds like some good applications in the warehouse. What about in the office?
Scott: In the office, I can see applications for voice-activated office equipment; I can see applications for really even just as an eavesdropping agent for any conversations between two people.
Jason: Please explain more Scott.
Scott: So, maybe eavesdropping isn’t the best word to use, but it can surface relevant data for a meeting for example, or maybe if it’s in a call center type of operation, you can have an assistant there that’s surfacing relevant information to that call for the operator. It’s just about being an assistant that’s there to help in those types of opportunities, but I know Alexa have an entire suite, Alexa for Business, something you might be able to shine some light on.
Nima: Yeah! It’s something that’s still pretty early days. It’s been run through the AWS part of Amazon and it’s basically, there are different functions you could use for different environments, anything as simple as like, in our buildings, not yet in Sydney, but in Seattle, you can walk into any conference room and say, ‘Alexa is this room available?’ Is there any of you that work in bigger offices and there’s always kind of that battle of like property?
There’s always some person that’s booked it for 8 weeks and never shows up, but you can book it in real time as soon as you just walk into the room or even starting a conference call. There’s a lot of exciting advancements that are around the corner in that space, but was just kind of simple but cool.
Jason: Regan, you talked a little bit about voice and optimising for voice, what are the things that we need to be doing differently when we’re optimising for voice?
Regan: Yes, I think that there’s three kinds of major things with voice search. Obviously local is a huge thing in voice, but there’s also answering questions and kind of understanding entities. So, your customers have questions and that’s become very prevalent even in the way people search on a desktop computer or a mobile, and expecting more from the search engine. So, I’ll say; ‘what is something?’ rather than just a few keywords into a page and I think that building out content for voice search and even for voice assistants to answers questions sets you up as the industry authority and people become trusting of you because you have the answers every time that they’ve got questions about that specific topic.
And I think the other application is–the major search engines are really trying to understand entities. So things like, movies and businesses, and the different parts that make up those, so you know, the actors or directors, and building up structured data through the various means is really helping to make this voice search experience a lot better when they can understand that information.
Jason: And how about the more transactional side? So I’ve heard the term voice commerce a couple of times in the office, if it’s not a term, it is now. How can voice be used in a more transactional environment?
Regan: Yeah, I think in the fast-moving consumer goods space, I think Domino’s is a perfect example of where convenience pays over price. If you’re able to just order your favourite pizza quickly, then you’re able to leverage yourself as a more convenient brand, rather than compete on price, which is always a race to the bottom. But I also think there’s some opportunities in the more micro conversions style and that stuff like I could imagine Mercedes, or something have an app which would let you book a test drive and do as micro conversions and because it has such less friction you do not having to fill in forms and whip out your phone and that type of thing, if you can be doing it in the car. Like I think it provides some opportunity to get to kind of really be there and provide less friction.
Jason: It sounds like there are lots of exciting themes and exciting opportunities that have been uncovered today. If I’m sitting in the audience, I’m thinking that this sounds very exciting, but it also sounds a little daunting and maybe quite expensive. The question I have for you, Scott is, how much of your existing web and business platforms can you leverage when you’re making the transition to voice?
Scott: Yeah, quite a lot. So, you just need to think of voice as just another type of interface into those systems. Those systems would already have the graphical user interface, typically a website, they probably already have the application programming interfaces or APIs for those systems to integrate with other systems. So really the work to enable voice, it all goes into the voice experience itself. So all that work is building something that uncovers the user’s intent, gathers the right information to service that intent and then it can call your API functions on your existing systems. So, there’s a lot of – you can leverage a lot of what you already have and focus all that new effort on providing a really good experience for the voice user.
Jason: It feels like, Nima, I was quite an early adopter of a voice, probably not as early as you, but it feels like it’s gotten a lot better really quickly. So, you know the some of the sort of early frustrations or challenges seem to disappear quite quickly and almost on a daily basis, the experience seems to be getting better. What’s driving them, what’s behind that kind of rapid improvement of voice?
Nima: It’s a combination of a few things, I mean the biggest advancements that have come in the last like 5-10 years, especially in ASR, which is Automatic Speech Recognition technology, is driven by advances in machine learning and cloud computing primarily, but I don’t want to give you guys overly simplistic view of it. It takes a lot of manpower and effort and energy and engineering talent to be able to make these systems not only work on just the base level Alexa experience or just voice experience, but for creating a seamless platform for any developer who wants to create a voice experience and have that just work. There’s quite a bit of thought and planning that went into that, but there is definitely some elements of machine learning that allow us to do this on the fly and like on an automated fashion and then being able to leverage incredibly powerful computers, which we have a lot of, it makes it a lot easier.
Jason: It’s always handy to have a couple of those. Scott, I might throw the next question, which is a bit of a curly one at you, which is obviously the topics of privacy and data security have been fairly present, you know, in our feeds and in our newspapers (if they still exist), fairly recently. Talk to us a little bit about what should be the considerations that organizations should be making when it comes to their sort of journey into voice.
Jason: Yeah, Okay!
Nima: The man on the street might be able to turn your lights on and off, but he’s not going to access your bank.
Jason: Nice correction! So, I think there’s probably a good moment to just open the questions up to the floor. I want to let people know that we also have people who are joining us via webcast, today, so, my understanding is literally tens of people [laughter] will be joining us. So, if they’re asking questions which kind of come from Johnny over in the corner there, he’s not a plant, he’s actually asking real questions.
Audience: [Inaudible 26:20] voice search you’ve already said 130% month-on-month. I’m just wondering if Google analytics actually provides that data?
Regan: So, I guess there’s a couple of things with the data, from search, it won’t really come through Google analytics, I think because most people don’t ever really match your site or get to your site, so the data’s going to be in Google search console because that’s where you see the impressions that your search results come showing up on a Google search. So the thing is, when Google for example, or Bing select a result, it’s usually the instant answer at the top of a result which just has the answer there and doesn’t require any click-through, so you’re going to get that data from Google search console because there is some metrics coming out of these major search engines and some predictions like Jason said earlier, ComScore is predicting that by 2020, 50% of searches on mobile are going to be voice search. So, it’s really on the rise and it is growing but yeah, I think there needs to be maybe a little bit more done to segment that out, but it should be considered a search just like any other, and so you need to keep that information in context.
Audience: Just want to follow on from the privacy discussion you had a minute ago, so obviously these devices are listening to you in your home all the time and are we to believe it’s only when they’re addressed and that they’re not harvesting everything you’re actually saying.
Nima: I hope you believe that, but if that is the case, the way that these devices are engineered and built from the beginning. Honestly, Amazon’s taking a pretty careful approach at this from the beginning and I can speak more to the skill side, than I can speak on the entire device architecture, but even the way that the skill work, like someone’s interacting with just say Dominos, we won’t even pass the exact text that the person is speaking over to Dominos and again third party. Amazon’s reputation with customers is pretty gold and precious to us, so we’re very protective of how that information works, so if someone wants to order a pepperoni pizza, the information that Dominos will get would be ‘reordering’ intent and ‘pepperoni pizza’ as values that are pre-determined in that application. So, we really break it down to its essential level and only pass along what’s important.
If your, you know, baby’s crying in the background or if your wife’s yelling at you at the same time, like none of that’s going to make it through,
The power of these devices is that all of the actual speech recognition is done in the cloud, cloud site, the reason that we’re able to make these powerful speech about, you know, voice assistance is because all the computing is done in the cloud, so we can’t do it on the device. The audio is sent to the cloud, but only parts that are used and sent to the developer are those intents.
Audience: So the rest is discarded?
Nima: It has a lifetime in this site or you can go in to discard that yourself if you want it to. Any of our users can do that.
Audience: So say you want to build an app onto Alexa what’s the actual process of it and how does the action work?
Nima: These guys [Inaudible 30:12]
Scott: It starts with defining what the use case is obviously, so, you need to have a think about the interface that you’re going to provide voice, so that involves coming up with some journey mapping just like you normally would in a graphical sense, but then it comes down to building a conversation tree and making sure that you’ve got some good happy paths in order to cover all the intents that you need to cover and make sure that you’re going to gather all that required information in order to service that intent. And then, you know, once you’ve got that sorted, then you need to build I guess a fulfilment layer, that responds, just that layer knows what the intent is, that it has information and then it can hit APIs on your systems to get the information or execute a certain function. Does that sum it up?
Nima: It’s an incredibly simple process to just play around with, I recommend all of you to just go and take a look at some of the one or two minute videos that kind of give you a quick overview. It’s a lot simpler than it was to build a mobile app. It’s as easy as deploying a web page and there’s tools that are coming, some that have launched in the US recently that will actually make it even easier for anyone to do it.
Audience: Hi guys! I’m sure you can appreciate that in order to get mass adoption for most of these new technologies, we need to have a seamless user experience, but unfortunately design is often an afterthought with tech lead innovation, I was wondering whether you had any advice for designers on how to approach it, who they should be talking to and what they should be telling their customers and their employees about this kind of thing?
Nima: Are you a designer?
Nima: You are the most important part of the equation. Especially and I’m being honest, that’s something that we learned pretty early on with the voice application use because developers, and I’m guilty this will just go and they’ll just whip something up and you quickly end up creating an IVR system where it sounds like a phone tree when you’re speaking to this voice skill. But we learned, when I say quickly and like after the first managed partner that we launched with, we learned that we need to have a design minded approach to each one of these experiences.
So, most of our time of all the development for the full lifecycle of a skill launch, about 50-60% of it is spent on design and then development is usually 20%, next is testing. So you really start with what it is that you’re trying to provide the user and go through that interaction back and forth, creating the conversational flow, and then finding all the areas where the user can kind of go off the path. If nobody says the same thing, nobody says something the same way, you know there are variations in speech patterns, in tone and pitch and just speaking style. So, there are a lot of tools that we offer that kind of help guide the user back and help the developers create proper error handling.
Audience: I’ve got a few questions if that’s okay. Basically, first question would be similar to the last question. Where do you actually begin, in terms of like, do you require a lot of programming knowledge to say start plugging into this service? Following on from that, what kind of security measures are there if you are linking into say you know ecommerce, account and things like that. Is there a chance to say, I want to change my order, or you know, whereas tracking orders and updating addresses and things like that, or adding things to your cart. What kind of security is there in terms of making payments and things like that via the voice?
Scott: Yeah! There are a few questions. So, in terms of how easy is it? I guess it’s like what Nima said, it’s relatively easy and you need to be comfortable being able to work with APIs, calling them and maybe building some of your own as well. And so, I’d say it not impossible, and the tools out there are really good and I think Amazon provide some great ones and there are some other tools, other platforms as well. In terms of the security side of things, yeah, obviously payments are possible but I’m not too sure on it.
Regan: So, you can integrate with other kind of authentication services, but to begin with you need to explicitly say that you give permission for the device to integrate with those, so it’s not automatically granted, so with Amazon pay or the Amazon account, I have to say share this information, so, you authorise that. With the payment services, from what I’ve seen, they look to a kind of authenticate either by something like a pin number or something along the lines of that to kind of provide security but I think you can opt into whether you want that or not. So, I guess the control is with you, but I guess with the payments, it comes down to, you’ve got to authorise those services to connect to the device in the beginning. So it’s up to you whether you want that, you can’t just pay somebody straight away, it’s not going to work, like you have to get permission.
Audience: Following on from that, how would you start educating your users to start using the service, like from a marketing point of view? If your customers have never heard of Alexa or Google Home or anything like that, how would you say, oh, now you can start using this? What would you recommend you sort of start educating how to ask the right questions?
Regan: I think there is power in building your brand, so if you’ve got a well-known brand, I guess utilising that is going to be really helpful. There is on some devices this thing called implicit invocation which is understanding the intent of what people want to do, and then providing a skill around that. So it’s not so much about the brand, that it’s about, you know, “order me a pizza.” So, like that invocation will go deep link into the skill or the action and it will create, will then offer up services generally that they’re able to fulfil that request and so, Domino’s, Pizza Hut or Pizza Capers or whatever. So I think that there’s work being done around that, but there is an education just thinking about what people would ask it on one hand and building a brand on the other, maybe even sharing it on mediums like social media.
Nima: I’m going to add two clarifying points, so the first two questions you had. First how easy it is to use. There’s a ton of just simple easy to use templates that you can just try out and you know, all the code is provided, it’s easy, you know, “insert joke here” or “insert question here” and deploy it, so it’s a good way to get a feel for what’s possible. Then the whole component with integrating with e-commerce – I think the way to think about this is you can integrate anything you want to, it’s all open platform that will access services that are available through an API. And So the security mechanism, first and foremost for any of these, these providers is a loss to, is what you would use to authenticate yourself as a user of that, let’s say of Dominos or Uber, that’s the first step.
And then some banks definitely that one extra layer of security on top of that for each interaction with the device. They’ll add voice pin as he was mentioning, or you know there’s others that they feel comfortable with letting you reordering a pizza after you’ve logged in with your username and password and that’s really more like a kind of a product and a policy decision for that partner. And we definitely have best practices, that now we got these partners on, but the point is that we can make any of those, you make any of these things, adding something to your card those things, I mean, whatever it is you can make possible, it’s only limited to your imagination and the skill and the will of your developers basically.
Jason: The sort of final question around educating users, have you got some examples that you’ve seen that seem to be that nice first step. So, you know you’ve just introduced your voice functionality to your environment, what are the marketing messages, you know, he’s giving examples of what a request may be, one of the kind of take some trips
Nima: Educating end users, right, like the actual customers? Honestly, I think that’s who we’ve gone through three iterations of our marketing emails that go out to partners or to customers and the one that seems to have planted, the one that’s most successful is one that just is a simple list of here’s the things you can say like, Alexa, you know, how many steps did I take today? Like, how did I sleep last night? So, you know, what can I cook for dinner? Alexa, play some Bon Jovi. There’s so much [laughter] every day for me. There is lot of what these devices can do that people don’t really know about. So, discovery is always a challenge and so helping people understand just what you can say and not overly complicating it, it’s probably the proven the most effective in my opinion.
Audience: Can I ask one last question please? So in terms of Alexa and Google home, how close are they in terms of what’s possible on both of them and are they pretty much on par with each other, or is Alexa miles ahead or where would you suggest starting with either both at the same time or would it be better to have one or the other?
Nima: I can be very cautious here.
Audience: I know it’s a risky question.
Nima: I feel that they’re both great at different things. I worked for a few years in the Amazon App store where we were the third place app store for quite some time. So, it was pretty exciting that we had such a leg up in this area of business technology. So, Amazon definitely has gotten a head start, but I don’t for a minute discount like, you know, the capabilities of the Google platform and what’s possible and honestly, I think that most people in Amazon were happy about Google stepping in the space, because it just validates voice and space in general. So, this is not like a fad that’s going to go away and this is how it’s going to, how things will progress and you can actually build for both and there’s tools that can help you do that and like, you know, deployed at each. There’s a lot more similarities than there are differences, but it’s just thinking about the end user from the perspective of voice is the best way to start.
Regan: I just want to add one thing, similar to designing like for screen sizes perhaps a different to look at where are your users. Surveying your customers, asking them what devices they have available to them because if they are majority Alexa, then you may as well start there.
Audience: Just keen to learn what apps you might recommend people start using this stuff with? So I’ve got an iPhone and that sort of stuff, never used the voice recognition, but you know, what would be kind of low-hanging fruit that you could start trialing with this as a user to start to become familiar with some of the technologies available.
Scott: So it’s some low barriers to entry into in terms of just start using it? So I think I touched on it earlier with what I use, is the simple stuff so checking the weather in the morning, you know, playing music
That through the weather app… No, that was like default behaviours.
Nima: How do you mean, like just with your phone, how you get started?
Audience: Yeah, this is generally so do you need the Alexa unit to use?
Nima: So, you have an iPhone or you can just download something called reverb which is, there’s two parts of the Alexa platform, there is one that allows you to build a skill and then there’s another one that lets you embed Alexa into any device that you want it to. So, it’s a free service that we opened and allow anyone to use and there’s actually a bunch of IOS developers that have created Alexa apps. They’re not managed or created by Amazon, but they’re just using our APIs. So, you could actually test almost a full Alexa experience for free using these apps, so if you just want to get a feel for what’s possible, how it works. There is also a web simulator you can use, it’s called echosign.io which was created by a third party developer that works the exact same way and you could go to the browser and use the full Alexa experience for free.
Regan: The devices are pretty cheap though, so you can get one easily reasonably…
Nima: He’s going to buy you one.
Regan: Kind of lower kind of ones like the echo dot stuff to get people started, which I think is pretty good.
Nima: Then you probably work, I mean after you play around with it for a while, it’s fun and especially just have something to have around the house.
Audience: I’ll stand – with regards to voice SEO, how does that work for brands and consumers in that with the screen, obviously you’re getting page one, 10 results, paid at the top, SEO and SEM and with voice you’re going to get what? How does that work and how do you get ranked as the top? What’s the business model for that?
Regan: So, I mean the kind of ranking remains similar to – to a part it remains similar, there are a little differences. So, it’s mainly focused around those instant answers or featured snippet boxes. So, there’s a bit of work around that you can do to optimize for that, providing schema and mobile page speed and the way you structure the content on the page, but essentially, it’s going take that snippet at the top of the page to return a voice result. So if I search for, you know, “who was the 44th president of the United States?” It will come up with the answer and it grabs that by kind of scraping your page and looking for the answer and if it’s usually in the first couple of paragraphs it kind of helps to provide that answer to people. I know say where it’s from as well, so it’ll say according to this website, “blah, blah blah.”
Audience: With a less binary answer “what is the best Thai restaurant near me?” for example, and I don’t like the first option, what do I do? Do I say, give me another option, give me another and does it just keep repeating that functionality?
Regan: Yeah, I believe so. Usually with most of these, it will come up with a couple of results. So, if you asked for what are the best Thai restaurants, I mean it’ll come up with a couple and it uses some information, like how much details are provided, the amount of reviews, what the star rating was, and contexts like that, so that’s about optimizing for local. But if you’ve got more information from Google or Bing to use, you’ll more likely to show up there, but also optimizing those listings around certain key words like Thai, I think is going to be able to help you boost that kind of local search ranking on voice.
Audience: Just interested to see whether or not you’ve got any figures on the actual penetration these devices. So obviously you’ve got Google home, you’ve got Alexa and in the U.S. it would be, you know, some kind of split and then how many year in, you know, how many houses? And then do you have any idea of what they consider plays in Australia? So, how many people are buying these devices? So if we create these things, how many people are listening?
Scott: What I read, I think it’s about, correct me if I’m wrong, but I think in the U.S. it’s about 20% of adults have a voice device, obviously not all Echos. Here in Australia, I’ve read that it’s currently kind of 4%, and it’s, you know, projected to again get up to the levels of around 20%, to meet the U.S. So, voice enabled in any device that you can well use the voice assistant.
Nima: We certainly don’t go into detail about numbers for these types of things, but for weight in, during one of our shareholder letters, the number that they shared, is tens of millions of Alexa devices sold worldwide. But if you did a simple search online for statistics around this, there are a lot of third party research groups that have done pretty detailed studies on what the projected numbers are, but that’s all I can say.
Audience: So just on that then, so we don’t know whether or not Google Homes are a major platform or Alexa in Australia yet?
Nima: It’s probably [crosstalk] pretty early on at this point. I think Alexa is definitely the larger player worldwide, but Google Home did launch here a bit earlier. But I think Australia is an interesting market and I think the adoption is going to be maybe a tiny bit slower, probably for a variety of reasons, but I think it’s just in a very early stage.
Audience: Just following on from that question, as an organization, if you want to provide voice search capabilities, you don’t want to exclude some of your customer base because they are on one ecosystem versus the other. So taking an approach where you can provide a consistent user experience across a couple of different ecosystems. How does that kind of work, I guess?
Scott: I think we’ve touched on it a bit earlier, but there are platforms that really focus on that voice user experience. It’s all about building the range of intents that you want to service, and all follow-up questions to gather all that information, and then from there that voice experience, that conversation tree can be deployed to a range of different platforms. It can be deployed to Amazon, it can be deployed to Google, deployed to chat bot interfaces as well. I think the tools out there to be technology agnostic and ecosystem agnostic are pretty good.
Audience: In terms of, if you want to launch total number of customers but you don’t necessarily know whether they actually have the physical devices themselves. Are you better off to create some sort of an app where you can just do it through your phone to start with, because again, you want to provide the functionality to a broader range of people without forcing them to go on and purchase hardware?
Scott: Yeah! Interesting! It’s an idea.
Regan: Some of these voice assistants are on your phone, so in Google’s example, they have the Google assistant built inside. If you’re going to build out something, you’re not only building out for that the device that you’re building it for the Google example, for the Google Assistant, which is on pretty much every android phone, but these platforms allow you to kind of push out the types of things as well. If you were to build it kind of technology agnostic, you could push out to a Facebook messenger chat bot or to even Cortana or Cisco Spark, any of those kind of Chat Bot or chat related interfaces where there is an interaction.
Nima: It’s not like building for IOS and Android where they’re completely different like work streams and efforts like 80% of the work that you’re going to do is going to be reusable in every form factor that you end up deploying this, and there are other tools that help you do it to your app or local app or you do it to through all the different platforms. But most of the work is reusable, so, it’s better to just focus on that part of it and then leverage these tools to get it everywhere.
Audience: I have a question here from Matt on Facebook. So, when was technology gets to a point where they can replace HR directors [Laughter] and you can tell me if that’s possible or not first.
Nima: It depends on what his HR Director is doing…
Audience: Is it possible still?
Scott: Should we ask Alexa?
Nima: I’ve had bad HR directors, I don’t really do anything. But my current one, they’re irreplaceable so.
Jason: [Inaudible 52:00] it just gets back. So, first of all, thank you, it was, you know, it’s always interesting when you have these panel sessions when you open up to the floor. There’s always that fear that you’ll just hear crickets in the room and things will come to an abrupt end. So well, it’s always a nervous experience when you don’t know what the questions are. I think it was really great that everyone was so engaged in the process.
So, we’ve now come to the end of our event and first of all, like to thank the panel. I’d particularly like to thank Nima for being here and sharing those insights and also, you know, being there for the, for the questions from the floor, which can sometimes be a little bit of a roller coaster for a brand, when you’re in the room, but, you know, I really appreciate the openness and the approach that you’ve taken today. And I’ve certainly learned something this morning. I’d like to thank from 4mation. I’d like to thank Julia, Claudia, Cassie, and April, for their work in bringing the event, together today and for Johnny for making it all work on the, on the day, except for his HR director question.
And finally, I’d like to thank you all for coming. My final thoughts are that voice represents, a rare opportunity for organisations to establish leadership in a space that’s becoming omnipresent and if you’re not doing so already, I’d encourage you to start having a conversation about voice. There’s been some really interesting questions raised and, you know, clearly we as organizations, we’re still very early in the journey and it goes without saying that 4mation would certainly like to be part of that journey.
So, to get the ball rolling, as attendees of the event today, we’d like to offer you a complimentary to our workshop here at our office, on the first hour will be focused on availability, and the second on utility, so if you’d like to take us up on that offer and book that session, please come and see me or any of the 4mation team. And again, thank you and have a great day.