It has been six months since we blogged about the development of Google's RankBrain and how it could impact SEO practice. As Steph advocated at the time, it is best to follow your usual practices and put creating high quality content and copy at the top of your hit list. Here's why.
In case you missed it or it's slipped from your memory, RankBrain is an AI Google ranking tool. It utilises machine learning to really nail down what it is that searchers are looking for with an ambiguous or wholly unique query.
It also, according to Google Senior Research Scientist, Greg Corrado, aims to be able to analyse pages on a sentence level and - even if they talk about the same topic in a different way - realise what they're about and rank them to a search query accordingly.
Understandably, there was a worry about how to optimise for this kind of programming.
In theory, RankBrain undermines the practice of keywording and looks to organise queries and pages in relation to context and natural language instead.
From Google's point of view, this helps their users who search in natural language rather than in keyword or organised question form.
E.g. "it wasn't Matt Damon but the guy who played an American spy in Europe." Rather than, "american actor spy film set in Europe 2000s" (consciously avoiding mentioning Matt Damon).
You can see why RankBrain is needed as these screenshots show that Google obviously homes in on the keyword "Matt Damon" and returns pages related to Bourne;
RankBrain aims to understand that the user is thinking of an actor like Matt Damon, who was in a similar type of film to Bourne (American agent, set in Europe etc. etc.), and return pages which talk about films which are of a similar ilk. But without users having to search tactically, like this;
This caused a worry because, when RankBrain reaches its full potential, you cannot accomodate what people are going to search for. In our analogy here, the searcher is thinking of some unnamed film and actor, but is openly including Matt Damon in their search query.
How could we account for that? Tactically mention every actor and film that might cause people to be reminded of ours? We can't.
That brings us back to Steph's original point - echoed since by the brilliant AJ Kohn article on Blind Five Year Old - the only way to cater for this is to create worthwhile, high quality content. We have to trust that RankBrain's contextual knowledge will grow to the point that it ranks our content correctly.
Spokespersons for Google have also stated, on record, that RankBrain isn't the leading ranking tool. It's more like third. And it's also used at the latter stages of page ranking.
Furthermore, the fact that RankBrain is a deep learning form of AI means that Google themselves don't know how it is going to grow. It will (they hope) learn to decipher unique, colloquial or over ambiguous searches and present the right pages to the user.
By learning which page the user decided best answered their query, it will rank it higher the next time somebody comes along with another unique variation of said query.
We can't account for this so keywords still reign supreme. Especially as voice searches become more common and users, conscious they are speaking to a device, choose their words more deliberately.
In short, the fact that Google don't control how RankBrain will sort unique/ambiguous/colloquial search results; it isn't the primary ranking factor; and it caters for such a small section of searches, means that it is nothing to fear.
Understanding RankBrain and realising you don't have to worry about it is one thing; but this is something you don't have to do. Get your SEO under control and start driving more traffic to your site, today, with these cheat sheet tricks.
Press the yellow DOWNLOAD NOW button to get out FREE ranking factors cheat sheet.