Virtual Assistant Optimization: The New SEO?
The SEO landscape has changed considerably in the last 18 months - the arrival of Google's RankBrain, specifically, has altered the process so much that now not even Google's engineers know, 100%, how results are being influenced. And now, there's another rising element to consider in the consumer discovery process, one that's only a small factor now, but which could become much more significant in 2017.
That new consideration is the rise of virtual assistants - and if the big tech companies have any say in it, these AI-driven helpers are going to become much more influential very soon.
Of course, virtual assistants are nothing new - Apple released Siri back in 2011 while Microsoft has Cortana built into their services. And while these are only early models of how virtual assistants can work, they've already lead to a significant shift. Google says that more than 20% of mobile search queries now come via voice search, which is projected to rise to 50% by 2020, while the biggest percentage of voice search users are teens, signaling a behavioral evolution. These trends are largely driven by assistant requests, and that's only going to increase as devices like Google Home and Amazon's Echo become more commonplace (and they will).
But while voice search is the future in private spaces, it's virtual assistants within messaging apps that are going to drive the next big shift in assistant tech for general use, and bridge the gap between the old and new search behavior.
And as such, this is where marketers need to start looking to ensure they're being found by people in need of their products and services.
Last year, Facebook revealed their Messenger virtual assistant project, called 'M'.
Only available in very limited capacity, M enables users to do more than merely ask questions and get answers.
As noted by Facebook Messenger chief David Marcus:
"Unlike other AI-based services in the market, M can actually complete tasks on your behalf. It can purchase items, get gifts delivered to your loved ones, book restaurants, travel arrangements, appointments and way more."
And while M itself hasn't expanded much beyond its initial test group, the learnings from M have been used in other areas.
Back in February, Facebook launched their new Bots for Messenger platform which enables brands to build their own automated bot systems, with these new AI response options informed, at least in part, by real world examples fed through the M system.
More recently, Facebook has been utilizing more such learnings, with new AI detection and response options being built into Messenger - back in September, Facebook started adding new auto-detect prompts in Messenger which use an AI system to 'read' the context of your message and offer options on actions you might want to take based on those engagements - for example, an in-stream option to make a payment to someone you owe money to.
This new capacity has been fairly low-key, but they are expanding it - as part of their new Messenger camera update, for example, they've introduced "programmatic frames" which, according to TechCrunch, can "extract meaning from common phrases and link them to different art". New tools are also being added to set event reminders and notifications based on the context of your Messenger exchanges.
All of these additions are progressing towards the next stage of the virtual assistant process, recognizing context in order to provide better, more useful options, which will eventually include in-stream business recommendations and options.
This process is currently best exemplified by Google's Google Assistant technology, which is built into their new Allo messaging app - within Allo, you can ask Google Assistant a question at any time within a message interaction and it'll give you Google search results relative to that query.
This is the next phase of such connection, and given the rise in messaging (text-messaging is the most popular form of daily interaction among teens by a big margin), it makes sense that the platforms are looking for smarter ways to tap into that behavior - even LinkedIn is developing a virtual assistant within their on-platform messaging process (and worth nothing, LinkedIn has seen 240% growth in messages sent on the platform since the launch of their re-vamped mobile app back in December 2015)
In time, all of these systems will be able to easily and quickly connect users to services, tailored to their individual behaviors and preferences - but what will that mean for businesses looking to be found by relevant searchers?
Virtual Assistant Optimization
As virtual assistants evolve, they're getting smarter, learning not only the basics of the search query entered, but also the surrounding context, along with non-communicated elements - i.e. your personal likes and dislikes, based on past experience.
For example, you might enter a query into Google Assistant like 'where should we go for dinner?' and Google will be able to show you all the local restaurants as it would display in Google search - those results are based on proximity and Google's regular ranking factors. But Google knows mobile searchers want more context - eventually, Google will likely refine their results to each user, referencing your past search history, and even location history, to understand more about what you like, then show you results accordingly.
This is even more of a consideration in Facebook's case - as Facebook CEO Mark Zuckerberg has noted in the past, Facebook is uniquely positioned to answer queries no one else can, based on the recommendations and actions of friends, whose opinions will carry more weight with the searcher.
This is the same impetus behind Facebook's recently launched 'Recommendations' tool, which enables users to post queries and use them as a call-out for response from those in their network.
Affinity is Facebook's potential advantage in search, that they know what the people you care about like, which is also likely to be of increased relevance to you - but given these more personalized responses, how can brands tap into these user flows and ensure they show up in relevant virtual assistant recommendations?
For Google, the key right now is ensuring your information is up to date and consistent to match with relevant location filters, but also, general SEO best practices should serve you well, at least until the provided results become increasingly personalized.
For Facebook, it's all about building presence and engaging with your Facebook communities - your business will stand a much better chance of showing up in relevant assistant queries in future Messenger interactions if you've been able to establish connection with your target audience, giving them some affinity connection to your Page.
Of course, both of these are easier said than done - ranking in Google and establishing a Facebook presence takes time and work - but in future, the more you can align to relevant communities and boost your digital connection with potential consumers, the better your chance will be of coming up as a relevant link in such interactions.
Either way, a strong online presence will become increasingly important, and considering new ways to establish brand affinity and engagement will go towards enhancing your digital standing and improving your relevance on both platforms.
Given the rising popularity of messaging, and the growth in in-stream, contextual search options, Virtual Assistant Optimization (VAO) is set to become a much more serious consideration in future. And while Google search will remain a crucial connective platform in this regard, it's important to also consider the potential of Facebook - with the vast majority of messaging users across both WhatsApp and Messenger - in this capacity, and the significant role The Social Network could eventually play in this new search process.
It may seem minor now, but as noted, the tech giants are placing big bets on messaging, bots and virtual assistants. It's worth considering the impact the likely growth of these services will have on the discovery process and where VAO will become a more significant factor, requiring specific focus, moving forward.
Follow Andrew Hutchinson on Twitter