Google Assistant Now Supports Over 1 Million Actions

Posted on January 9, 2018 by Paul Thurrott in Cloud, Mobile, iOS, Android, Smart Home with 33 Comments

With Google Assistant heading a breath-taking number of devices this year, the search giant is highlighting another key strength: This technology now supports over one million actions, which are essentially things the Assistant can do.

“To start the year off right, we’re at the Consumer Electronics Show in Las Vegas showcasing the Assistant at home, on the go and in the car—and all the ways it can help in each of those places.,” Google’s Brad Abrams explains. “It’s been an exciting year to see the platform expand to new languages and devices and to see what [third-party developers have] created.

Developers that create apps for Google Assistant build actions into these apps, expanding the capabilities of the Assistant and making it more useful for users. These actions can span a range of wide capabilities from language translation to online purchasing to games for kids and much more. Over time, as these actions become more and more sophisticated, Google Assistant—like Amazon’s Alexa—becomes smarter, and enables more conversational experiences.

(We’re still waiting on routines, I believe; this feature will let users string multiple actions into a script-like entity that will enable you to say things like “Hey Google, goodnight,” and have the Assistant trigger multiple actions in one go.)

Of course, with over one million actions already available on Google Assistant, just finding out what the Assistant can do can be a bit daunting. So Google has created a new directory on the web, and an updated directory on mobile, so that both developers and end users can learn more. (On mobile, open the Assistant app and navigate to Explore.)

This ecosystem is exploding: Google notes that there are over 400 million Assistant-enabled devices in the world. And that Assistant works with over 1,500 smart devices from over 200 brands, with more on the way.


Tagged with ,

Join the discussion!


Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Comments (33)

33 responses to “Google Assistant Now Supports Over 1 Million Actions”

  1. arunphilip

    A million.

    And Cortana supports just 300 (and dropping):

    • david.thunderbird

      In reply to arunphilip:
      I see a lot of negative but it is from those sitting on their asses staring at their empty glasses (Cortana) instead of getting off their asses and filling their glasses (Cortana) with tasty beverages (APPS). Out of the 7 billion people on the Earth those that can use APPs each one may only use five to seven but they will have thousands of good ones to choose from for their pleasure like OK Google users.

    • jrickel96

      In reply to arunphilip:

      A million. How many are worthless or will never be used? How many commands will people memorize?

      We've run into this before with command lines. Most people are not interested in having to remember a ton of commands, so having millions is pointless. You focus on the core of your audience.

      Many of those million options for Google will be going away at some point too as support is ended by many participants in the ecosystem due to low usage. I already know of several larger firms that are quite well know and have marketed their Google Assistant and Alexa integrations that are reconsidering because the numbers are terrible for usage.

      These assistants will make phone calls for you, schedule something on a calendar, and play your music for you. They will lose most of their advanced functionality that is dependent on support from third parties because that functionality is already becoming so watered down in an ecosystem that is very tiny. There's only 10s of millions of real potential users right now and active users are much smaller. Phones use these assistants primarily in cars, but usage rates are much lower than touchscreen usage for things like Car Play and Android Auto - people prefer the touch screen to voice commands.

      Google and Amazon are using their marketing might to tell you that this is the future. They're lying. The adoption rate is the worst for any major digital tech in the last 30 years. PC growth, smartphone growth, and tablet growth all happened at a much more furious pace.

      Hand it to Google and Amazon for selling it and convincing people that it is the next big thing when it likely isn't. They've suckered a lot of companies into supporting them - but I know several companies are furious behind the scenes because the growth projections and monthly usage rates they were given by both Google and Amazon were far, far from the truth.

      So be warned - there's a backlash coming from companies that have supported these digital assistant skills. Google and Amazon both have painted the rosiest possible pictures of growth, yet it is not even 10% of what they had projected. In a few months, I believe the general public will begin hearing about this and we may see the digital assistant market and that smart speaker market start to decline by the end of the year.

      Remember when Siri was the next big things and everyone used it for awhile on their phones - and then usage died down. Alexa and Google Assistant on speakers and elsewhere are novelties now - but using them is not organic.

      • hrlngrv

        In reply to jrickel96:

        . . . You focus on the core of your audience. . . .

        Is there just one homogeneous audience? Or could there be several different though related audiences?

        If some of those actions were related to sports, couldn't each sport have its own actions? If I only care about baseball, I'd only need to learn the N actions specific to it, and all but scores and injured players would be specific to baseball and useless, say, for ice hockey.

        Likewise actions for finding the nearest gas station with the lowest price if I didn't own a car, or anything related to public transportation for people living in rural areas.

        Far less like the command line than like music or TV show preferences.

        • jrickel96

          In reply to hrlngrv:

          Trying to be all things to all people often makes you good at nothing or very few things.

          Amazon and Google are only out to improve themselves with these assistants. Neither company really makes much of anything. Yes, Amazon does a little, but the bulk of their money is made by selling stuff others make. Google sells ads. That's the only thing they do that makes them money.

          So both their assistants are about increasing those ends for them. They want to add volume so they can help themselves out and they won't care about how well anyone else on their platform does. That's how it is with Android. Very few make money on Play, but they have to be there. That's on the case with these assistants nor is it likely to be the case any time soon.

          • hrlngrv

            In reply to jrickel96:

            I grant that Amazon and Google have little or no interest in selling their own-branded physical products, maybe not even their own-branded services. However, if one wants an eventual customer base numbering in the billions, just how focused can one be?

            Another example: if one type of action were buying movie tickets, wouldn't there need to be different actions (so generating higher action count) for each country if not each city if not each theater chain in each city?

            My point here is that it matters critically how actions are defined. Given the presumed ad value of being able to claim MANY actions, I suspect action is defined rather narrowly, metaphorically in the sense that closing one's right hand into a fist is a different action than closing one's left hand into a fist.

      • Stooks

        In reply to jrickel96:

        "Remember when Siri was the next big things and everyone used it for awhile on their phones - and then usage died down."

        I bet if we had actual numbers from all players that Siri would but either at the top or near it in terms of usage, simply because it is on so many iPhones in use.

        I will be the first to admit Siri is not the "smartest" right now but I use all the time from small things like "Call soinso" or "timer 20min" or "Play" some music.

        Amazon is the hot new item but it is not on any phones as of yet.

      • arunphilip

        In reply to jrickel96:

        I bought that argument of quality over quantity when I bought my Lumia some years ago. It makes sense up to a point, but no further. And when the Uber app on Windows Mobile (for example) worked far less optimally than the equivalent app on Android/iOS, that quality argument fell flat.

        Bear in mind that every user is not going to use the same 300 skills, or benefit from all the 1 million skills. However, the odds of users finding relevant skills are higher with a pool of 1 million to choose from.

        I've not yet got an assistant baked into a speaker. I'm waiting to see how the dust settles between Google and Amazon, and until then I have Cortana on my PC and Google Assistant + Cortana on my phone to keep me happy.

        • jrickel96

          In reply to arunphilip:

          I got a Google Home last year. Haven't used it since last January. The speaker isn't great, so I don't want to listen to music on it. There were a few cool things that were interesting.

          But why would I want to do a search with it? I don't want a Wikipedia article read to me. Asking about the weather is a nice feature. Asking for a news report is a good thing.

          I you have some Hue lighting, it works well to command that.

          But I'm not sure the daily function will extend beyond that for most people and for most it will not warrant spending money on an ecosystem they use for very little.

      • kevin rose

        In reply to jrickel96:

        It doesn't matter that there are a million things. What matters is that amongst them there will be one, two or maybe three, that will be invaluable. I already see people using their phones to navigate, to find products, to control light and heating in their homes.

        I am absolutely sure that digital assistants will be a success

        • offTheRecord

          In reply to GrumpyOldGit:

          Yes, this is similar to the issue with apps on smartphones. Obviously, no one person needs a million "actions," but not all users use the same actions. Some users will find certain actions indispensable, and others will find other actions indispensable. For many, there will always be those one or two actions you just can't do without (just like with apps on phones). Any platform that doesn't have those one or two actions is going to be a non-starter. An "AI" platform with a million (and growing) actions is probably going to better meet my needs now and in the future than a platform with "only" 300 (and shrinking?).

        • jrickel96

          In reply to GrumpyOldGit:

          I'm certain they'll be a huge failure. Some of the AI from them will be useful for other interfaces, but they are novelty. Usage and adoption rates are historically bad.

          The NBA can only manage a couple hundred active monthly users on Alexa when they draw millions of eye balls per game and have millions of people that use their app on their phones, Xboxes, etc.

          Alexa's top skill is playing Jeopardy. Novelty.

      • Nicholas Kathrein

        In reply to jrickel96:

        I agree but disagree. Yes there are many unless things in that million but there are more useful ones there too because there is a million total. I agree that at the moment you have to learn just what to say to use them and so they won't get used but as the AI gets better it will take those million of useful and non useful skills and just take any words you say and be able to figure out what you want and which skill does it and boom you have it working just by normal talking. Does that take 1 year or 5? I don't know but I think Google has a big lead here from that aspect and will be the 1st to be able to do that thing I described and that's when this goes from a speaker that sets timers and plays music to an actual helpful thing that you place everywhere in the home that just does things for you.

        • chrisrut

          In reply to Nicholas_Kathrein:

          Agreed. It was several years after the microcomputer was invented before that infinitely useful thing called a spreadsheet converted that technician's toy into an indispensable tool.

          Just last night my brother-in-law (a professional German translator) and I were discussing the limitations of translation - even for humans. Without going into the details, the bottom line was that some things can only be translated in their complete context.

          Likewise, those "million things" an AI can do (and the millions to come) only achieve true usefulness when the AI can figure out when and how to use those abilities (and crucially important, sets of abilities) in the context of our lives, guided by policy we set.

          BTW, I'm a huge supporter of voice-activated and autonomous-driving technology in autos, because anything that makes the driver look away from the road is a bad thing...

        • jrickel96

          In reply to Nicholas_Kathrein:

          I think the digital assistants will change a lot and the "let me talk to my devices" stuff will never become a mainstream thing. I also expect that if Amazon cannot seen increases in shipped goods and Google does not get increased information, that both with EOL their digital assistants down the line.

          One advantage MS has is they have a clear purpose behind Cortana that's not just about gathering data, but they want to do something useful to the user.

      • VancouverNinja

        In reply to jrickel96:

        The novelty of Alexa Echos has worn off in my household - we have three of them in place and almost zero use now. It would be the exact same with any of the others.

        • jrickel96

          In reply to VancouverNinja:

          That's generally what the analytics data says. Usage rates are now under 30% of estimated installs when it comes to tracking monthly active users - and many of those people may only be asking to listen to music or for a weather report and nothing else.

          Usage rates are also declining, not growing. There's a small subset that does use these things regularly, but it's very small.

  2. Bats

    I just bought a new car, a Mercedes Benz. It had the option of either installing the Mercedes multimedia package or the Android Auto/Apple Carplay package. FYI, the latter package allows you to install BOTH Android Auto AND Apple Carplay. What did I choose? I wanted to choose the Android Auto, but I opted for the Mercedes multimedia package because it had a ton more cool features, plus, I could always hook my Pixel 2 XL to the car's system and "android auto" that way.

    However, that's not the point of my post.

    The point of my post is that Mercedes can be used for both Google Assistant and Amazon Alexa, where you can literally tell either digital assistants to perfrom functions, such as start the car, lock the doors, etc.... 


    I setup the Google Assistant to work with the Mercedez Me "Action" and....boy, oh's AWESOME with a capital "A." The Google Assistant acts like a third party, where you (the user) ask her to tell Mercedes Me to perform the action you want. For example "Hey Google, tell Mercedes Me to start my engine." People, when I performed this action and went to check my car....OMG, it worked! Here is another thing, you can even command The Google Assistant (via Google Home) to talk to Mercedes Me. When you do that, you hear another female voice come from the Google Home speaker and you can just say commands to her.People, this is sooooooooooooo COOL! It's supercool! In this cold tundra called the Northeastern USA, 1) I no longer have to physically go to my car to start it to warm it up, and 2) I don't even need to go to my phone to use the Mercedes App to remote start the engine.

    All in all, I think it's safe to say that whenever I buy any new tech, whether it's an appliance or not, it's gonna have to be compatible with the Google Assistant (or Alexa). It just works. 

    Also, unlike Microsoft, Google want this space. Google said we are now going to be an "AI" company. Boy did they mean it. All this AI stuff is great, I encourage everyone to buy this great tech and enjoy......before they revolt and kill us all (lol).

    • puggsly

      In reply to Bats:
      My concern with this is the number of hands in the pie! You talk to your phone through the google app which transmits your voice securely over the internet to google's cloud which talks back through the internet to your phone to communicate through another app to a Mercedes interface on the web which then communicates back through the internet to your vehicle to start your car.
      This allows for great flexibility but both google and Mercedes knows specifically about your car and this is the way for most "skills" based IOT systems work.
      HomeKit's method makes more sense even if it will take longer to support everything.
      I talk to my phone which does also pass my voice to Apple and it does return the text to my phone but that ends their specific involvement. The text is processed locally on my phone and the HomeKit routes the command directly to my device ether via my local network connection or if I'm remote by making a secure connection to my network and controlling it directly via my network.
      This may seem like a little thing but no outside service knows about any of my IOT devices and at least in theory, that includes Apple. This has to greatly reduce my visibility on the internet and thus my risk.

    • Stooks

      In reply to Bats:

      My 2014 Chevy Truck has this button on the remote. You push it and the truck starts. OMG!!!

      The action takes probably 1/8th the time it takes you to do the same with Google Assistant if not less. Google or no one else knows that I did this. It works from anywhere in my house. It works from my 3rd floor office at work to the parking lot below. OMG!

      Apple Car play in my Truck is pretty amazing as well with my iPhone, nice and private as well.

    • VancouverNinja

      In reply to Bats:

      Really? You need to ask you car to start? I push my pedal....some people push a button....

      You are excited to tell an AI to talk to another AI instead of doing it directly...These are two examples of useless features that take more time to use than doing it the normal way.

      You can use google Assistant to warm up your car remotely...every vehicle that does this today provides remote on/off features.

      All you have done is reinforce that Google Assistant is more useless features than being helpful. I am sticking with Cortana thank you. It actually does things that I had not been able to do easily before.

      • kevin rose

        In reply to VancouverNinja:

        I disagree. I used to work in the middle east and the ability to remotely turn on the a/c in my car would have been great. I'm starting to use Google more and more. I am mobile, my phone is mobile. Cortana just does not offer anything to me. I guess it's a lifestyle thing.

      • chrisrut

        In reply to VancouverNinja:

        I disagree. Any technology that can prevent the need for drivers to look away from the road to do things is justified on that basis alone. It will save lives and money. That there are stupid, pointless ways to use the technology is both obvious and irrelevant - would you suggest people throw away their phones because there are dimwit telemarketers out there?

  3. Michael

    >>>Google notes that there are over 400 million Assistant-enabled devices in the world. And that Assistant works with over 1,500 smart devices from over 200 brands, with more on the way.<<<

    How does this compare with Alexa?

  4. Mark from CO

    Dear Microsoft:

    This is how you develop, introduce, and support new consumer products. If you are really committed to the consumer space, as your leaders have said, you will have to be able to do these things, but even better.

    Please also note the ancillary benefits - Google is developing many relationships with larger companies. Don't be naïve and think Google won't use these relationships to help their enterprise initiatives.

    Mark from CO

    • hrlngrv

      In reply to Mark from CO:

      You believe everything MSFT senior managers say?

      Perhaps actions (or should that be INactions) speak louder than words. Perhaps the dearth of MSFT's smart device partnerships is the most accurate measure of MSFT's interest.

      • Mark from CO

        In reply to hrlngrv:

        hrIngrv: I agree totally. Perhaps I was not sarcastic enough. I do find it amazing that a senior Microsoft manager will say and believe (I accept he believed what he said) that Microsoft is pursuing a consumer presence, when reality is sooooo much difference. Are Microsoft leaders that delusional?

        Mark from CO

  5. Lateef Alabi-Oki

    Didn't Paul write an article last year about how Android Auto and Android TV were failures?


    Anyway, it's nothing short of impressive how Google in just 3 years is looking to be the player most likely to dominate the living room and our homes.

    Today Google sells more variety of hardware for the home than Apple, Microsoft, or Amazon. They also have a larger and far-reaching platform and ecosystem in Android, the most dominant computing platform on earth.

    And they have the most powerful virtual assistant to tie their hardware efforts to their AI prowess in ways that are both unique and compelling.

    This year's CES just confirmed what I already knew last year, Google, not Amazon nor Apple, remains the company to beat the not so distant future where computing is ambient. And if they play their cards right, they will become the platform for Ambient computing.

    Amazon has to be given a lot of credit though, they brought out in Google what no other company has been able to do. They've forced Google, for the first time ever, to compete in a way I've never associated with the company. And I think it's made Google a better company in general.

    Let's be honest, the Google Assistant wouldn't have evolved as fast as it has done in the last year without Amazon.

  6. Waethorn

    Articles like this seem to bring out the tech conservatives.

    I think the problem you see with these is that most normal people don't like using this stuff because they don't want their tech to be thought of as their friend anymore than they want to have a conversation with their toaster (watch the Red Dwarf episode about this for a laugh). I think people want to perceive tech as being something rudimentary that does something when you press a button, not a "brain" predicting what they want (even though AI isn't a real thing yet).

    • chrisrut

      In reply to Waethorn:

      I can share the sentiment, but not the conclusion; let me explain (and sorry if this is too pedantic for some - - it's my field :-).

      Technology is that which extends humans beyond their inherent limits, along any vector. The urge to do this is so hard-wired in our psyches it is broadly (and I believe correctly) considered a - and perhaps the - primary distinction between humans and other creatures. We don't wait for genetic evolution to "grow a new finger" if we need one; we go to the shop and make one (leveraging the tools of intellect and an opposable thumb which nature provided).

      In that context, I don't agree that people - most people - want technology to be rudimentary. They want it to provide easily understood advantages over their native capabilities, and they want the "costs" to be less than that perceived value of those advantages.

      Along with money, those costs include ease of use in the broadest sense. The "handle" of the tool - i.e. the UI - must require as little effort as possible to both learn and use. For a very obvious example: the handle of a hammer should be free of "splinters" (in the UI sense, think Windows 8... :-)

      Having said that, I believe any reluctance to adopt AI will stem more from reluctance to adopt the unfamiliar, and pay the real or imagined costs, than from a desire that it be rudimentary. Consider this: the computer technology that has exploded across our planet in an eye-blink of history is anything but rudimentary. Yet, when micro-computers were first emerging from the labs, a banker said to me "Why would anyone want a computer on their desk?" The advantages were not obvious, and the costs of acquiring and using those primitive devices was high.

      This rubric sets the bar for evaluating new technologies: the difference between a hula hoop - a fad - and a tool, is that tools fundamentally extend our capabilities, while fads amuse us by their novelty. But novelty soon wears off; they don't fundamentally change our relationship with the environment, as tools do.

      To anyone interested in the topic I recommend two books: Diffusion of Innovations by Rogers, and Understanding Media (part 1 in particular) by McLuhan.

  7. daveevad

    "Google Assistant Now Supports Over 1 Million Actions"...And 999,999 of them involve stealing and selling your information :P

  8. wright_is

    But can I get it to start podcasts and audio books?

  9. melinau

    Most of them are hard to use, don't work very well and completely useless for what I actually want to do... But apart from that its great!