Google RankBrain is Applied to All Search Queries, But What is It?
To sort through search results, Google uses a machine-learning artificial intelligence system called “RankBrain”. RankBrain is responsible for processing qualitative written data into quantitative mathematical data. Having to go through a ridiculous amount of written content every day is no good if it cannot be understood by computers. By processing the data into vectors, the algorithm can process the search results. As part of the wider Google algorithm, RankBrain does not function alone. It is actually a specific part of the Hummingbird algorithm. Hummingbird has been around a while, but RankBrain is one of the newest parts of the overall search algorithm.
The Third Most Important Ranking Factor
According to Google, RankBrain is the third most important ranking factor. Coming in right after content and links. It has major importance to ranking, and there’s a chance you may not have even heard of it! Ranking factors are determined by hundreds of signals that go into the Google algorithm. With RankBrain being the third most important signal out of hundreds, it’s definitely worth paying attention.
Specifically, RankBrain is incredibly effective at handling never before seen search queries. By guessing what words or phrases might have a similar meaning to the query, the system can make an informed guess and filter the results accordingly. This is why RankBrain has such an effect on SERPs – it is responsible for filtering and ordering results in order of relevancy. According to Moz “it is believed that the query now goes through an ‘interpretation model’.” This interpretation model analyses possible factors affecting the query. For example; the searcher’s location, personalization, and the words of the query. Important signals like user location and how new content is must be taken into account in order to interpret intent and deliver the results most likely to satisfy searchers. The end goal is to determine the searcher’s true intent. By understanding this true intent, Google can deliver the most relevant results.
Optimise for RankBrain
Preparing your content for RankBrain is neither new nor complicated. Instead, focus on the structure of your content, optimised images, and an engaging and positive user experience. Unlike other algorithm updates, there are no penalties or guidelines to follow. As content for RankBrain goes through an interpretation model, it cannot be specifically optimised for. Something that does matter, however, is context. Making sure you capture that context is a key to optimizing for this machine learning approach.
Search Engine Watch said “As RankBrain learns on the job, it will only get better at understanding semantic and concepts. It will also improve at understanding relationships between topics and queries. This will benefit voice search result accuracy as well as traditional search results pages.”