Founder and CEO, Labworks.io, Tom Hewitson joins us to share his advice on how to build a business on Alexa and Google Assistant. read more
If you’re considering using voice assistants like Alexa and Google Assistant for marketing, here’s a few things to watch out for. read more
Join three of the voice AI industry’s leading minds to debate whether voice is living up to its promise and what should happen next. read more
A good digital assistant will take context into consideration when providing a user experience.
Now that context can be related to the device that you’re using, could be related to the environment that you’re in, could be related to how much time and attention you have available at any given time.
So for example, if I’m in the kitchen washing up, I might have a bit of time but you might not have my attention and so the experience might need to be different to if I’m sitting in the front room watching the TV, where I do have time and I do have attention or if I’m out for a run wearing headphones and I don’t have either and so in the headphone example, maybe your interactions need to be really short and sharp and transient. In the living room example maybe you use visuals a little bit more and you lean on visuals more and in the kitchen, maybe you use audio first and you try and emphasize using earcons and things like that to make more of an audible experience.
Now, those are just real high-level examples and it’s difficult enough to create one conversation that’s intuitive. That’s natural. That’s easy to use.
Now think about doing that for all of these different devices and think about doing that not just for one third party app that you create but if you are the designers behind Google Assistant, it exists on over a billion devices, in over 90 countries and 30 different languages.
How do you create conversations that, yes, adapt to the different devices that you create as Google, but also the any number of devices that could be created by third-party manufacturers putting Google Assistant in their own hardware.
That is a very complex, very big task but it has to be the task for someone, and that someone is Daniel Padgett, Head of Conversation Design at Google.
He and his team work on creating consistent conversations across modalities for Google Assistant and we had the opportunity to interview Daniel and chat multi modal design for Google Assistant on the VUX World podcast this week.
We talked to Daniel about just how you go about creating genuine multimodal conversations that change depending on the device and context the user is in and where the future of multimodality is going from Google’s perspective.
Google’s Head of Conversation Design, Daniel Padgett, shares how his team approach multi modal design across all Google Assistant-enabled devices.
Jon Bloom, Senior Conversation Designer at Google, joins us to share what a conversation designer does at Google, as well as some conversation design techniques used at Google, such as ‘grounding strategies’. read more
Voice search is happening. And Google is under threat. Not short term threat but long term threat.
There’s a new search provider in town and it’s name is Alexa. read more
At Project Voice in Chattanooga, TN, we saw some light shed on just how many people are using the Google Assistant, and some updated numbers on the kind of traffic one of the most popular Alexa skills handles.
Cathy Pearl, Head of Conversation Design Outreach, Google, towards the end of her presentation shared some numbers on how many monthly active users Google Assistant has.
500 million per month.
Thats across all surfaces and devices, but it’s a lot.
Google is the first of the big players to share numbers like this. Maybe it was waiting for the number to be high enough to warrant talking about.
I’d like to see the same numbers from Amazon. Now I’m wondering whether Amazon is also waiting to reach a noteworthy milestone before releasing them.
300,000 per day.
2% of users try the premium service and 90% of those convert into paying customers.
That’s doubled since we spoke to Nick on the VUX World podcast last year.
Of those 300,000, 15% of them were using a smart display of some kind, such as the Echo Show 10, 8 or 5.
So now that we’re starting to see some solid numbers coming from the platforms and smiths successful developers on those platforms, it would be good to see others follow suit and share the metrics behind their successes in order to validate the opportunity that these platforms present for organisations and individuals.
If you’ve been uncertain about the kind of opportunities that exist on these platforms for your company, or cautious about jumping in and investing, maybe these numbers will start to help you paint a clearer picture about whether you should and the size or the market and opportunity that you could take if you do.
Here’s a prediction for you: Google Assistant will replace google search entirely within 5 years.
Think about it. What is Google’s ultimate aim? To find THE right result for the user.
That’s why it’s taken on the burden itself of finding hotel rooms, flights, deals and just about anything else it can right there in the search engine results pages.
It’s consumed all facts (more or less) within its knowledge base and it’s even pulling website content into the search results pages in featured snippets to save users the trouble of visiting the site.
Now Google Assistant will replace the voice search on Google search.
And what does a voice interface mean for search results?
Result Zero wins. The holy grail. One result. The best result for your needs based on your context.
Google Assistant is also text-based. So why not replace the search bar entirely with a conversational AI? One where search refinements aren’t refinements at all, but clarifications:
“What year was Rob Williams born?”
“Did you mean actor Robin Williams or singer Robbie Williams?
That’s not a search refinement, it’s a clarification statement.
Google is able to take your first search term and turn the interaction into a turn in a conversation.
When you have a follow up query, there’s no need to head back to the search bar and delete what you’ve just typed, you can just continue:
“And when did he die?”
Google has been working at integrating the web into Google Assistant for a while starting with the beta launch of speakable markup. This allows web editors to mark up specific sections of content to be read by Google Assistant.
Then look at the last I/O ‘19. Google released the ability to turn YouTube videos into tutorials on the Assistant and web-based ‘How to’ tutorials, too. And featured snippets appear in Google Assistant all the time.
It doesn’t stop there
It goes deeper than that. In-app actions mean that not only can Google Assistant serve website content, but it can also pre-populate Android apps with spoken or typed phrases from Google Assistant.
If you say:
“Hey Google, book me a small hire car for Tuesday”
In-app actions will allow Google Assistant to pre-populate the Enterprise Car Hire app with ‘small car’ and pre-select ‘Tuesday, all day’ as the date.
Once this has been rolled out, it’s obvious that the next step is to work on an integration that goes the other way: from the app, into Google Assistant.
This way, you’ll be able to book your car, hotel, flight, cinema ticket or do anything else you use apps for, right there in Google Assistant, without needing to open the app at all. It’ll just use the app’s APIs and a conversational layer that you’ll be able to include in your apps through a more robust in-app actions feature.
At I/O’19, Google changed its mission from ‘organising the world’s information’ to ‘helping people get jobs done’ and in-app actions aren’t part of that.
So, once you have in-app action integrations working both ways, the APIs and language models can be broken out of the apps and surfaced through Google Assistant on the web, which will add reliable capability to ‘get jobs done’, right from within the most popular website on the planet.
So my prediction, again: Google Assistant will be the front end to the entirety of Google Search within 5 years.
Imagine that. The words information condensed into a simple conversation. All of the Android app functionality available instantly over the web using the same APIs.
So if you have a website or an app, regardless of what your business does, you should consider whether to brace or embrace.
Brace yourself to be forgotten or embrace voice first and jump into Google Assistant this year.
When thinking about launching an app for a voice assistant, most people, most of the time, will create an Alexa skill instead of a Google Assistant action. But why?
There are over 100,000 Alexa skills and only a few thousand Google Assistant actions.
Now, Google do claim that Assistant has over 1 million actions, and that’s because Google treats its definition of action slightly different to how Amazon defines a skill. But in terms of third party crafted experiences, comparable to Alexa skills, the Alexa skill store dwarfs Google Assistant’s.
Why aren’t people building more Google Assistant actions?
People aren’t creating actions because they don’t think that there are enough users to warrant the investment…yet.
This is largely because of the misunderstanding among some people that voice assistants and smart speakers are interchangeable words.
People think that voice is smart speakers and smart speakers are voice.
Perhaps that’s where it started in 2016. But voice is used more often on phones than anything else. Siri and Google Assistant are the two most used voice assistants, despite Amazon’s smart speaker market share dominance.
History repeats itself
What’s playing out right now is the same story that played out when Apple launched the App Store.
When the App Store launched in 2008, everyone started building apps for iOS.
Then Android released the Android Marketplace and… Nothing.
No one was building apps for Android. But why?
Simple. All the users who were using apps were iPhone users. Yes, the Android Marketplace was established in 2008, too, but Android didn’t overtake iOS market share until 2012: four years after the launch of the app store.
Even then, not everyone turned straight to Android. By that point, people had invested thousands of pounds into their iOS apps. They were sitting tight for a while until it became unavoidable.
Google is in the same position again
Through no fault of its own, Google has managed to find itself in the same position with Google Assistant as it was with Android in 2009. It was second to market with the smart speaker, second to launch a developer toolkit and second to launch its actions ‘app store’.
And although Google Assistant is available on over 1 billion devices and is running on the no.1 phone operating system in the world, as long as people associate voice assistants with smart speakers, it’ll continue to struggle.
Google’s task at hand
Google’s task is the same as anyone else trying to compete with Amazon, or anyone else working in the voice industry: helping people understand that voice is more than smart speakers. Voice assistants can exist on any device; mobile, ear buds, car stereos, washing machines, computers, laptops, TVs and, yes, smart speakers.
That’s why it’s leading with the ‘1 billion users’ slogan. It’s trying to give people confidence that Google Assistant is a safe place to invest your time and energy.
And this is a message that they have to hit home because they’re going to struggle to claw back market conceded share on smart speakers. Baidu and Xiaomi are already gaining traction and don’t be surprised if we see more entrants into the market in 2020 from the likes of Sonos who now have Snips’ technology.
So people aren’t creating actions because of a lack of understanding about what Google Assistant actually is and where it exists, which should give Google a clear focus for its marketing and developer/brand relations with Google Assistant in 2020.
This post was triggered by a discussion on the Voicebot 2019 year in review podcast. It’s the answer I wish someone gave 🙂