Did you know that every time you conduct a Google search your query gets processed by the search giant’s artificial intelligence system known as RankBrain?
Back in 2015, RankBrain was only being used for around 15% of searches conducted via the Google platform. Today, it is used for every search – approximately 2 trillion per year – and it’s changing the rankings of lots of queries.
Your average Google user has probably never heard of RankBrain – let alone understands its role in searches – but according to Google’s plan and currently available data, the decision to process more through the AI system will make the web an easier, more responsive, more accurate place to search.
So What Is RankBrain?
RankBrain is an artificial intelligence system, rooted in machine learning, which is applied to and used with Google’s current search algorithms to provide a better experience for users. It combines mathematical processes and an advanced understanding of language semantics to learn about how and why people search the way they do. It can then apply that learning to future searches to make them even more accurate and rewarding for the user.
The real beauty of RankBrain is that it is always learning and can update itself in real-time, rather than being pre-programmed and having to rely on human input to get better – like many other aspects of Google’s search function. In short, RankBrain owes it’s success to the rapid evolution of artificial intelligence.
RankBrain And Ranking Signals
RankBrain is actually now considered a new ranking signal and a very effective one at that. In fact, of the hundreds of “signals” Google uses when it calculates rankings (a signal could be the user’s geographical location or whether the title of a page matches a search string), RankBrain is now rated as the third most useful behind content and links.
This is a very important observation, in that there are hundreds, if not thousands, of ranking signals used by Google every time a search is conducted. For RankBrain to be considered the third most important, highlights its significance in the overall search puzzle.
How Is RankBrain Improving Search Results?
It’s important to know that RankBrain isn’t actually changing the rankings of search results for every query, but rather for “a lot” of them, as stated by Google.
Think of RankBrain as more of a query refinement tool than anything else.
For example, a Google user may search for “best restaurant in Vancouver.” RankBrain is designed to realize that the search string is similar to other more popular ones, such as “best Vancouver restaurants.” As a result, RankBrain will choose to display the results of the more popular search rather than the results for the user’s actual search, thus improving the relevance of the results and, therefore, the user experience.
If you’re a search marketer, RankBrain shouldn’t be something you’re worried about. Despite it handling every search request, there’s nothing you can do other than focus on creating great content, great keyword strategies and aligning both to every web asset you own.
That’s because even people at Google have stated that they don’t know how RankBrain does what it does – though how true that is can be taken with a pinch of salt! After all, as CEO, Sundar Pichai, recently stated… “be it search, ads, YouTube, or Play, you will see us — in a systematic way — apply machine learning in all these areas.”
Surely this would not be the case if they ‘didn’t know how RankBrain does it’.
For the masses, our focus should remain on great content – whatever the digital platform on which it is being displayed. This has and should always be the case. Following that principle, RankBrain should reward great content accordingly.
For more on RankBrain, how it works and what it means to digital marketers take a look at a great video from Rand Fishkin at Moz: