Google has said on numerous occasions that they use over 200 ranking signals, but lately there’s a lot of talk about RankBrain. It is the third most important ranking signal after links and content, but there is a lot of speculation on what it’s all about and how Google uses it for ranking. The fact that RankBrain determines rankings based on machine learning and to a certain extent that’s considered to be Artificial Intelligence (AI) either fascinates or intimidates. But it’s just a single component of the Google’s ranking system which depends on more than 200 rankings signals.
All of the ranking signals Google uses could be based on machine learning, although that may not happen any time soon. The reason why the entire Google core algorithm cannot be based on machine learning is that it would become difficult or almost impossible to do any debugging. It wouldn’t be easy to debug machine learning or AI decisions as there would be too many layers of neural networks. Gary Illyes has stated so in numerous articles. So let’s focus on RankBrain and what it means for SEO.
Rank Brain- Rise Of The Machines
It’s important to remember that Panda and Penguin are also based on machine learning systems. Google is using machine learning to come up with new signals or to aggregate signals. That helps them to keep improving the search results they deliver to users. RankBrain uses historical signals to do its re-ranking of results from Google’s core algorithm. This type of ranking filter started applying in October 2015 and now contributes to all Google search rankings. Machine learning allows Google ranking signals like RankBrain to keep improving the overall ranking system over time. The improvements happen more gradually with machine learning than they do with human controlled ranking processes. When humans apply changes to the algorithm, the changes take place all at once and sites are affected in a big way. So as ranking signals based on machine learning systems become smarter and smarter, there will be less of unwanted surprises and the best SEO strategies will get rewarded almost on real-time basis.
RankBrain plays an important role within Google’s search ranking mechanism as it quickly learns from historical data and refines the search results especially for search queries. For complex search queries using long-tail keywords, Google confirmed that tests performed showed that RankBrain does better than its own engineers. It is therefore expected to be more helpful to experienced searchers that tend to look for more specific answers using long search queries. Machine learning in that case helps the search algorithm to best understand context, semantic relationships and even need for personalized search results. Check out our new JSON LD Tool to help build your semantic profile.
RankBrain as an advanced ranking signal based on machine learning makes it possible for Google to deliver a great experience for searchers. The search results they get largely depend on the meanings of their search queries rather than the exact words they entered. So SEO Agencies will be forced to pay little attention to keywords, but to focus on making sure content is as helpful to users as possible. Understanding the user intent will be critical in order to make the content great for them. With the Google’s core algorithm expected to become more dependent on machine learning and artificial intelligence going into the future, it would soon become almost impossible to game search rankings if that isn’t already the case.
At some point, Google may just achieve the goal of a search algorithm that quickly learns form search queries, and automatically updates itself to improve on the accuracy of search results without human intervention.