RankBrain is part of Google’s comprehensive algorithm used to sort through billions of pages it has information on to find the ones most relevant to a search query. It is not a new way Google uses to rank pages. Rather, it is a new addition to the existing ranking algorithm – Hummingbird.
Google considers its RankBrain to be the third most important ranking factor it uses to process search results and rank websites. Launched in 2015, RankBrain is a machine-learning artificial intelligence system, and below we’d like to share everything you need to know about it to make your page as recognizable as possible.
#1: Understanding RankBrain
Before we go any further, we’d like to take a look at what ‘machine-learning’ actually means to better understand RankBrain. ‘Machine-learning’ means that a computer is teaching itself how to perform a task instead of being taught by humans or by following a computer software. When it comes to RankBrain, it looks as if ‘machine-learning’ and ‘artificial intelligence’ pretty much mean the same thing – a system designed to teach itself and make connections.
In addition to RankBrain, Google also uses PageRank, another part of Hummingbird, to rank websites based on the links they use. Another ranking tool Google uses are signals on a webpage, namely more than 200 major signals, and possibly up to 10,000 sub-signals, while RankBrain is their third instrument or signal that determines how a website will be ranked on the search results page.
#2: What Does RankBrain Do?
RankBrain is used to provide search results when a page doesn’t contain the exact words that were used in the search. It can connect things that may at first appear unconnected, see patterns, and understand similarities that might not be that obvious. Once it figures out these patterns, it can associate groups of searches with results that it thinks Google users will find the most helpful. RankBrain is good at deciphering ambiguous searches and producing great answers to queries.
In other words, it’s concerned with what you as the user were thinking when you typed in your keywords. Before RankBrain, Google had trouble figuring out what you actually meant when you typed in ‘grey console developed by Sony’. It was simply searching for pages that contained the words ‘grey’, ‘console’, ‘developed’, and ‘Sony’, as Backlinko beautifully explains. But now Google knows you actually mean ‘PlayStation’! It strives to understand your thought process much like a friend would, which is pretty amazing stuff!
Further, RankBrain is hard at work offline too because that’s where it does all of its learning. It is given old searches and it learns to make connections based on these. The connections are tested, and if it passes, RankBrain goes online for us to use.
#3: What is RankBrain’s Biggest Concern?
Like any good friend out there, RankBrain is most concerned with our (user) satisfaction. It turns user keywords into concepts, sits back, and observes how the user responds to its search results. It gives a set of results and checks out how you interact with it. If you’re not satisfied, it will produce a different page for you next time, reducing the Google ranking of the initial page. If you’re pleased with how RankBrain did its job, it will uprank the page you chose. When observing your reaction, RankBrain looks at your click-through-rate, dwell time, bounce rate, and pogo-sticking.
Wondering what pogo-sticking means? Well, let’s say you searched for ‘best used cars on the market’. Like most people, you click on the first website in the results page only to realize its introduction is tedious, it’s informing you why you should get a used instead of a new car (like you didn’t already know that), and it isn’t telling you which models to look at. So you go back to the results page after a few seconds. You click on the second page, and no luck again. You go back to the results page, click on the third suggestion, and bingo! You finally find the information you are looking for, stay on the page for a good several minutes, and don’t go back to the results page after. This going back and forth is called ‘pogo-sticking’, and RankBrain pays very close attention to it when assessing your satisfaction with its work.
#4: What Do I Do With This Information?
Well, you either tackle your webpage optimization yourself, and pray you’ve got a good grip on all the SEO tools that the pros use to increase any page’s ranking in Google. Or, now that you know how important every single aspect of SEO optimization is to your visibility, you hire a team of professionals and let them do their magic as you sit back and admire their genius and that of Google’s algorithms to recognize your website as that worth ranking on page 1!