What Is Rankbrain?

  • By Siobhan Snaith
  • 21 Sep, 2017

Google is the top search engine on the planet and handles more than 40,000 queries a second, which is an average of 3.5 billion queries per day. This is a big task to execute, and they use the help of a machine-learning artificial intelligence system called RankBrain to process and select the required data.

Artificial Intelligence

The developers at Google had to delve into detailed programming in order to build RankBrain, which uses artificial intelligence to refine the matching pages before returning the most relevant search results.

Artificial intelligence enables a computer to be equally as smart as a human, in the respect that it acquires mathematical and grammatical knowledge through building on obtained information to create new connections.

It’s not the determining factor of how search results rank, but it is part of the algorithm Google calls, Hummingbird.

Google Ranking Signals

Google uses different signals to rank web and blog pages, and they run into the excess of over 200 ranking signals. RankBrain quickly rose to be the third-most important ranking signal months after its deployment. That brings about the question, how does RankBrain have an effect on returning results?


Its primary purpose is in the interpretation of queries to find pages that have words not entered in the query but still closely related. Before RankBrain came about, Google would not display the results of ‘jogging’ if you tried to search for ‘jog’ even though they are variations of each other. It helped to make Google smarter by incorporating synonyms, for example, when you search for court shoes, you could also mean formal shoes.

It also brought understanding that there are searches for different demographics such as a search for BlackBerry could mean the phone manufacturer or the fruit. With the huge number of searches Google receives every day, RankBrain helps to interpret the search queries better while effectively translating them into ways that Hummingbird will return the most relevant web pages as the results.


Since its inclusion as a ranking signal in 2015, RankBrain has made a significant impact by constantly making small tests then launching big changes when it has a huge degree of confidence. Google would not do it unless it was necessary. Programmers at Google say that RankBrain learns to predict the best possible and relevant web pages by using historical searches

Share by: