Introducing your very own personal Google assistant who knows everything about you!

The Internet of things (IoT) has long been a buzz word in the digital world, however much of the transformation that we have seen until now has predominantly come from Microsoft. At the SMX West’17 conference today we heard from Google’s Jason Douglas about how Google Assistant is going to transform search by being always available, always accessible and always knowing where you are, what you like and what you want.

Google says that they have spent a number of years refining the new Google Assistant technology which will better enable a conversational experience between a user and Google to get things done in their world, across a range of connected devices – whether at home, in front of the TV, or in the traffic on their mobile device. In an inter-connected world, users should be able to find what they want and get things done with ease, no matter where they are or what device they are using.

SMX_Google_Assitant

What is Google Assistant?

As with any good assistant, the Google Assistant needs to know you almost as well as you know yourself. It should know what you want and what you like all the time while you’re on the go, no matter where you are, or on which device you are connected. Google Assistant should serve you in the same way that a real-life human Personal Assistant would, provided the Assistant was with you 24 hours a day. You should simply be able to ask your Assistant a question naturally having an ordinary conversation in your own language, and your Assistant should be able to easily interpret your every need and desire. In addition, your Assistant should know what kinds of food or movies you like, which devices you are using and where you are at the time of your request.

Right now, there are many things that Google knows about you and your world, and their intention is to make it easier for you – not only to organise and find the world’s information – but now also to complete everyday transactions and tasks, like boiling an egg and timing it just right!

Critical components to the Google Assistant Efficiency

For your Google Assistant to be able to solve any problem you might have, whether it’s running an egg timer or finding out what time you can expect your package to arrive that you ordered online, there are a number of components that Google Assistant needs to perfect in order to be effective:

  1. Context: Knowing what you like so that it can match your preferences to your desires.
  2. Content: Understanding your calendar and whereabouts so that it can ensure you always arrive on time and have what you need to be most effective.
  3. Tasks: Knowing what you want to do and how you can best complete a particular task or transaction.
  4. Connect: Knowing which device you are on so that the assistant can always be accessible to you whenever you need it, no matter where you are.
  5. Intent: Understanding exactly what your intent is without you having to repeat yourself on every device.
  6. Identity: If you have already logged in, your information should be passed across to all devices seamlessly.
  7. Transactions: Your payment information should be securely stored across all devices in order to make it as natural and easy as possible to pay for transactions, while being able to receive after-sale fulfillment (such as knowing when your package will arrive) with ease.

The evolution of Google Assistant

Google has been working on refining and perfecting the Assistant over the past few of years, with the launch of Google’s Knowledge Graph (understanding the relationships between people, places and things); the introduction of the new Hummingbird Algorithm (understanding natural language voice search); RankBrain (artificial intelligence and machine learning); and Google’s ongoing commitment to perfecting Google Maps. Google has built up a substantial amount of information about you and how people just like you look for information to fulfil their needs.

Having gathered all of this information and refined its technology, Google says that is the right time for them to start rolling out Google Assistant, however that it is still early days and a lot is still being done in the background to perfect the technology in order to ensure that it does not give the user a bad experience.

With access to all of this information across all of these devices, there is sure to be a lot of concern about ‘big brother’ watching your every move and storing all of your personally identifiable information. Google says that they have and always will take your privacy very seriously and that it is imperative for the Assistant’s survival that it be trusted; and that they will continue to work hard to retain this trust. There is always consent involved when sharing any personally identifiable information, and Google will always give users as much control and transparency over their privacy settings as they always have.

Is there going to be shift from the Mobile First index to a Conversation Action first index?

Greg Sterling (Search Engine Land) commented at SMX that about 60% of users who are using voice search do not expect to see the same results as a conventional result. Does this meant that we are going to see a shift from optimising for the mobile first index, to a new voice first index? Google commented that ranking signals for actions will be very different to ranking signals for mobile first, however it is still too early to say whether Google will switch, as they are still learning about user actions and needs. However, they did say that success in Assistant’s results will mostly rely on the inventory a website has on its site and which services the website (and the user) have used in the past. As mobile and voice may continue to diverge, there are many new ranking challenges that still need to be figured out as the cost of a bad transactional result certainly raises the stakes.

How can websites become more discoverable on Google Assistant and what does this mean for SEO?

In typical Google style, they did not divulge any ranking secrets, however websites developers should most certainly continue to design websites that provide conversational experiences with users. Considering that task completions and transactions is the overall objective that the user will use the Google Assistant for, a conventional search engine result may not always assist the user to fulfil this request, and this is what the Google team is working on to leverage the experience on search so as to assist the user to completely fulfil a task or transaction with its help.

While the basic considerations and design principles are fundamentally different between conventional search results and transactional or actionable results, Google is still learning and developing best practices towards streamlining the process. Google will be adding more capabilities over time and will start to roll it out to various platforms more aggressively. However, since the Knowledge Graph has been a core component to the development of Google Assistant, applying semantic and structured data mark-up, and describing intent on the website will no doubt become very important.

How to build websites for conversational experiences?

Here are a few key takeaways to consider when building for Google Assistant:

  • Be more targeted with messaging
  • Apply efficient structured data mark-up
  • Create more conversational content
  • Ensure your websites are mobile-friendly
  • Make sure pages load fast
  • Consider Accelerated Mobile Pages (AMP)

In closing, when creating content for your website, think about the identity, intent and context of the user that you are trying to connect with. Build out detailed personas and map relevant and natural conversational content to address their problems through meaningful conversational experiences.

Google Assistant Actions

Google has released Conversation Actions guidelines (https://developers.google.com/actions/) to developers to assist them in building websites for the Google Assistant. Conversation Actions help websites better fulfill user requests by enabling two-way dialog with users. When users request an action, the Google Assistant processes this request, determines the best action to invoke, and invokes a Conversation Action if relevant. From there, the action manages the rest, including how users are greeted, how to fulfill the user’s request, and how the conversation ends.

Contact us for assistant on optimising your website for Google Assistant.
Author: Gillian Meier, CEO of BlueMagnet