On 10/26/15 Google announced RankBrain, a machine learning artificial intelligence system that helps them serve better search results. Ready to learn about RankBrain and SEO! Let’s do it.
Read the ultimate guide to RankBrain (just launched on Ignite Visibility).
What is RankBrain?
RankBrain is a Google ranking factor as well as an artificial learning machine that allows them to decipher the best search results. RankBrain is a component of the Google algorithm, not the entire algorithm. Using artificial intelligence, this technology can aggregate information concerning user’s intent and satisfaction when searching on Google and use that information to produce better search results.
How Does RankBrain Work With Google SEO?
At this point, we only know what Google has told us. In addition, we can piece together information from historical Google news.
RankBrain is part of the overall Google algorithm known as Hummingbird. We know this is true because in the Bloomberg report it states that RankBrain only handles a fraction of the Google search results. We also know that Hummingbird has other components. It looks something like this:
Hummingbird Algorithm
Each of these items is a component that handles different parts of the Google website. What is interesting, is that Google has said the following from the Bloomberg article:
“RankBrain is one of the “hundreds” of signals that go into an algorithm that determines what results appear on a Google search page and where they are ranked,” Corrado said. “In the few months it has been deployed, RankBrain has become the third-most important signal contributing to the result of a search query.”
As we can see, RankBrain has already been out for a few months and it is the third most important signal! So that left everyone asking; what are the first and second most important signals? Google of course declined to answer, but my guess is links and content.
What Does RankBrain Do?
Danny Sullivan over at Search Engine Land was able to get Google to answer a few questions via email. He states that RankBrain looks for content that matches search queries, even though the content might not have the exact keywords. Now, the issue is that Google already did have a way to match a search for say something like:
Best Baseball Hitters in San Diego
With something like:
A List of the Best Batters in the Area
This was actually a big change we saw after Hummingbird was introduced. Google started providing much more variety in the search results. They also didn’t seem to be showing as much exact match results. Finally, we know that Hummingbird really started connecting things together based on schema markup.
The Details on RankBrain
RankBrain started to gradually roll-out in early 2015. It seems it has been more full steam ahead since September 2015. Google has not told us exactly how many queries are impacted, but they did say a very large fraction.
Google said that RankBrain actually does all of its work on servers and is not a live Internet-based system. Instead, they feed it historical search information. It then takes that and makes predictions. It then takes those predictions and tests them. When they work, they are pushed live. Next, they simply repeat the process.
So all of this makes it seem like RankBrain is just an aggregator of some kind, however, Google has actually confirmed that there is a component of it that determines if a page ranks or not, making it a ranking factor.
Google has stated that if you want to learn more about RankBrain, you need to know how words and phrases are connected. Google is able to match language together mathematically. You can read more about it here.
Also, there is a very complete research paper on the topic here by Google.
Summing Up RankBrain and SEO for Now
This information is pretty new to us, but it is certainly important. We have known that Google is connecting similar works and data points for a long time. Even back in 2010, most SEO professionals knew what latent sematic match was.
Will RankBrain change SEO? Yes! Do we know how much? No! Do have new strategies yet? Not yet! Will we soon? Absolutely, already working on it right now actually…
