Google�s machine learning system, �RankBrain� is designed to help process its search results. That makes understanding how RankBrain works essential for SEO success.

The terms artificial intelligence (AI) and machine learning (ML) are often used interchangeably but they are two different things:

Artificial intelligence can be described as the way machines operate intelligently. It�s the broader of the two concepts, encompassing machine learning and other �smart� tasks or operations.

Machine learning is a subset of AI that refers to how machines process data and learn on their own.

As technology has changed, AI has changed as well. Where the field used to focus on programming machines to complete complex tasks and calculations, the rise of the internet changed things. Now, with access to massive amounts of data, machines can be constructed to think like humans and make decisions on their own, as opposed to following instructions that told them what to do.

This affects search by allowing algorithms to try and understand the content they come across to deliver better results to search users. AI is one reason the �old hat� ways of optimizing sites no longer work.

Google is putting AI and ML to good use. RankBrain is the machine learning process by which the Hummingbird algorithm works. RankBrain allows Hummingbird to sort through multitudes of data to find content that most accurately reflects the search query.

It is particularly effective on �long-tail� queries and unique queries because it is designed to recognize patterns and connect them with seemingly unconnected search queries. It alleviates situations where computers would typically get stuck on queries they were not familiar with or did not understand.

RankBrain works in conjunction with other factors or signals like links, content, and mobile-friendliness to help Hummingbird rank websites based on search queries. It refines search queries by identifying synonyms and related searches to return more appropriate results for users.

For example, if you searched for �shoes,� RankBrain would know to also include websites that contained the words �sneakers� or �flats.� As queries are conducted, the information goes into Google�s database of indexed pages, and RankBrain is able to learn which content is similar and which is not. It then uses that information to deliver more appropriate results as other queries come up.

In this way, RankBrain makes it less important for SEOs to use the exact right keyword to reach their intended audience and more important for SEOs to provide context around the keyword using the appropriate content.

« Back to Glossary Index