Now more than ever the SEO industry looks to be mirroring that of politics, with sound bites and buzzwords the order of the day.
Grand lanyard-heavy conferences, emphatic speeches thrown across eager audiences, the sharing of likeminded ideals, and the spinning of content — it’s all there.
I’m not even talking about SEO: The Movie, where link building meets Wolf of Wall Street meets the losing players of The Social Network.
I’m talking about SEO and artificial intelligence.
Over the past few months I’ve seen a large amount of respected agencies talk about offering AIO specialisms to their clients. That’s Artificial Intelligence Optimisation to you and me.
But there’s a snag about offering AIO to your clients, because AIO doesn’t exist.
At least, not in the way that’s currently being presented.
Although many agencies talk about Google’s RankBrain as a form of AI that can be manipulated, the truth is that it just isn’t. Google’s RankBrain algorithm is an example of Machine Learning, which by definition is not the same as artificial intelligence.
By definition, RankBrain is closer akin to a form of Artificial Narrow Intelligence (ANI). It’s a non-sentient intelligence focused on a specific task, much like Alexa and Siri.
The intelligence comes from its ability to “learn” based on a number of other variables, a gross over simplification of this would be understanding that someone searching for “Mexican Food” may find results about “Tacos” useful.
RankBrain is tasked to learn the relationships between these things to give the user a “smarter” search experience.
It’s an algorithm specifically designed to help process search queries and embed written language into vectors so that computers can understand. It sits alongside the other algorithms, under the umbrella of Hummingbird.
RankBrain is just one of hundreds of ranking signals, which then have their own variations and sub-signals (so one of potentially thousands).
In the now infamous Bloomberg article, Google says that RankBrain signals carry the third most weight in search results, but given the number of signals, this weight can’t be insurmountable.
And while it’s still in such an infantile stage, there’s a lot of speculation about how the algorithm will interact with the internet in the long-term.
One thing that we’re sure of however, is that you just can’t optimise for RankBrain. This comes from Gary Illyes, a trends analyst at Google in Switzerland, who said that there is no RankBrain score.
When it comes to voice search, RankBrain uses other algorithms and semantics to understand user intent.
Search histories, user proclivities, and other factors are used by the algorithm to understand where the context might belong.
It also takes into account backlink profiles and compares similar websites to establish regularities alongside testing for breaches in Google’s link scheme guidelines.
This is another reason why websites need to stay within their own spheres of influence and operation, so that a backlink profile looks as natural as it should be.
If it finds backlinks that it associates with spam in your given sector, it’ll throw your site in the spam bucket too.
As for products of RankBrain, such as Rich Snippet, Google has given absolutely no guidance on how to deal with them.
The team within Google responsible for Rich Snippets and Rich Cards, has point blank refused to write any sort of guide for the subject.
In recent Google Webmaster Central Hangouts John Mueller, also of Google Switzerland, has stated that RankBrain is “flowing into our algorithms in a general, but not in a very targeted way at the moment”.
At the end of the day
Dealing with RankBrain isn’t the great specialism of the age that agencies are beginning to make it out to be. Artificial Intelligence isn’t a new concept, Bing introduced RankNet back in 2005, although like much of Bing’s algorithms, it hasn’t been overly-publicized.
In days past it was a standard tactic for agencies to stuff keywords and farm links to get results.
Today we work on multifaceted campaigns that encompasses technical, content, link building, and security.
And in the future we’ll be looking to cater to algorithms just like RankBrain, the ones that take away the cheap loopholes and add more technical into the equation, much like the Fred quality update in early 2017, and Possum in 2016.
Given the updates made towards ensuring web pages deliver quality (and aren’t focused purely on getting a conversion), as well as moves towards making the user experience safer, it’s hard to see these practices becoming old hat in the near future.
Like before it’s about catering to an audience and ensuring that you are providing users with the right kind of information that they’re searching for.
It’s about strategic high quality writing and search intent, which is exactly what agencies should have been focussing on for years.
According to many, many well-respected SEOs (and a large number of other blogs), optimising for RankBrain comes down to 3 main areas — fresh content, creating a high E-A-T and brand authority online, encouraging engagement onsite (I’m not talking about “bounce rate” et al), and creating a strong orchestrated internal linking structure.
This allows users to navigate between useful pieces of content and webpages that satisfy their queries.
Advertising AIO seems to me a little like advertising that you’re finally doing your job, which begs to question just what you were doing beforehand.