I’ve heard RankBrain called a number of things since it was announced to much hyperventilation towards the end of last year; an algorithm, a ranking factor (3rd most important if you believe some sources), artificial intelligence and a robot squirrel called Dave. Only three of those are true.
Actually, RankBrain is kind of an algorithm but more accurately it’s a machine learning system.
What does that mean?
Back in the day (which in the fast moving world of digital marketing can often mean yesterday) we would talk about LSI (latent semantic indexing), a theoretical part of G’s search algorithm that would look not only at the words used in a document but also at the surrounding words, and the words commonly found on other documents with similar topic relevancy to getter a better handle on searcher intent. For example, if I search for ‘ball’ am I looking for a football? A tennis ball? A black tie dinner and dance? It might not be until some smart programming has analysed the context and the other words that correlate most highly with ‘ball’ across the entire web that a better guess can be made as to what I’m after.
Of course, that by itself is not enough to fathom the kind of ‘ball’ I’m after. I may be looking for a used kabadi ball from the 1987 world kabadi championships (I might!). In this case my search history might come into play, my geographical location or perhaps my social media activity might lay some clues as to my hobbies and interests. Answering difficult, highly personal and often ambiguous questions requires a smarter cookie. Enter RankBrain.
It’s not artificial intelligence but it is aiming to be artificial narrow intelligence. This means a machine learning system that can be better than human intelligence at a given task; like being a search engine for example.
Honestly we don’t know an awful lot about it, as with all of Google’s projects there is much postulation, gossip and panic abound, but what we do know is that it is real and it is currently active globally. It is working on your search results right now.
Hardcore programming theory is only going to send us both to sleep so let’s bottom line this in terms of SEO and your digital marketing strategies.
What is isn’t and what it is going to do (probably)
Like Hummingbird, mobile-ageddon, panda and others before it, RankBrain is not going to destroy you. It’s not seeking to filter anybody out and penalise them. It is attempting to make results more accurate for increasingly more complex search queries. It is trying to gauge searcher intent and deliver more accurate search results.
It is going to reward the long tail; all those thousands of different search terms that send traffic to your website (rather than the one or two terms that we used to obsess over – you don’t still do you?) are going to become more viable results for Google to send back to searchers.
All that high quality, well written, unique blog content that you’ve been working on to support your social media marketing and long tail SEO strategies are going to be even more critical to future proofing your positions and search visibility.
You don’t really need to optimise your content for RankBrain, you simply need to be aware of it’s existence and that smarter algorithms are only going to become even better at analysing and organising content across the web, not just for relevancy but in terms of context and intent.
For me it really points us directly at a future where outdated search factors like keyword density, link count and content length are finally buried and consistently high quality content will stand alone as the lynch pin to all your domain strength building, social engagement winning and general traffic growth goals.