Google: If I Only Had a RankBrain
December 28, 2015
Image courtesy of: Keegan Berry
When you and I are shopping or looking for something on the internet, usually there’s a specific purpose behind it. Some of these could be “where can I find X” or “how do I find out more about X”, or even “I want to purchase X”. And it would be in everyone’s best interest if we could get to our desired location, or target information in as few steps as possible, because let’s face it: we’re all in a really big hurry!
The problem with this is that people aren’t always sure about the best way to reach their desired information on the internet, and the search queries that they use when browsing the web become extremely ubiquitous because of this.
The Need for improved Semantic Search
To address this, Google and other search partners have been working on better ways to bring the desired information to us, instead of vice-versa. This effort is known as Semantic Search. The Wikipedia definition is this:
“Seeking to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable dataspace, whether on the Web or within a closed system, to generate more relevant results.”
This concept is exactly what Google (now under Alphabet) has been striving for since they joined the search game. And they have been able to stay at the top of the search totem pole for quite a while now by coming up with new algorithms, carousels, penquins, and new gadgets that help their users get there quicker and easier.
Along Comes a RankBrain
A year ago, we were talking in a BrightHaus blog post about where Google’s A.I. interests would be going in the future, and we have recently been getting a slight glimpse of this. Google has recently unveiled their latest toy in the mission to improve semantic search, and it is a form of A.I. software known currently as RankBrain.
Greg Corrado, a senior research scientist at Google spilled the “artificial beans” in a Bloomberg article recently about RankBrain, and provided all of us SEO’s some insight about what the new phenomenon is all about.
Similar to the concept from terrifying Will Smith sci-fi movies and Isaac Asimov novels, RankBrain is a form of A.I., or artificial intelligence, that utilizes a basic form of machine learning and helps Google and their teams define, measure and implement their precious ranking factors. Previous to this, humans have been solely behind Google’s ever-changing functions of processing all of the information on the internet, determining the relevance to possible search terms that people might use, recognizing patterns and connectors, and then relaying that information back to all of us “Googlers”. These humans have been doing a fairly good job at recognizing and predicting trends for Google’s algorithms, but the problem is that it takes humans a long time to do this effectively, the results end up inconsistent, and are by no means perfect. And they have rolled RankBrain out to help them analyze this information not only quicker, but with better accuracy.
RankBrain Put to the Test!
According to the Bloomberg report, some of the top Google Search Engineers collectively went head-to-head with RankBrain in order to determine exactly how accurate it really was. They were all given a bunch of information from the internet and were tasked with ranking it based on relevance and search intent. The end results were even better than Google had even anticipated! Not only did RankBrain process the information a lot quicker, but it even predicted the information search relevance about 10% better than the human Search Engineers!
So, RankBrain has been considered a big success for Google, and has already been working feverishly for them. In fact, Greg Corrado admitted that RankBrain is now possibly the #3 ranking factor for Google SERP’s. According to a recent article in Moz, There are currently hundreds of ranking factors that Google uses to determine search relevance. And being revealed as the #3 Factor is a big deal because Google won’t even share what their #2 and #1 top ranking factors are at all. They are Top Secret.
Even though the phrase “ranking factor” has been used for RankBrain in many publications so far, there are some skeptics, such as in this article on SearchEngineLand where they speculate that it’s more of a query processor than a ranking factor. Only time will tell exactly what RankBrain truly does, but what we do know right now is that it’s a pretty major part of the Google system.
How Will RankBrain Affect Search Engine Optimization?
So SEO’s and online marketers alike are scratching their heads and trying to think what this will mean for the future of SEO? One can only speculate at this point, but I tend to agree with the assessment that it’s heading SEO into a great direction, and that Google will reward those SEO’s that have adopted good, white-hat SEO practices that revolve around the creation of great content and excellent user experience. For example, an article on Kissmetrics’s Blog, similarly mentioned that SEO will depend more on high-quality content, engaging writing and usefulness, and less on freshness, link-building, and technical SEO.
Because with the aid of RankBrain and some miraculous intel, the Google systems are becoming collectively smarter, and are beginning to understand the connections between keyword variations and definitions. For example, the Google algorithm is starting to better understand not only that the queries “shoe” and “shoes” go together, but also that synonyms like “shoes” and “sneakers” should be paired together as well. And you can assume that they are going to take this one step further and preemptively fill in keyword gaps, and even link sequential searches together, so if the first search query is “green shoes”, and the second query after this is “green ones in size 11”, Google will soon begin to understand that the user is talking about “shoes”.
So one can assume that the practice of pulling endless lists of keyword variations, pluralizations, misspellings and synonyms and stuffing them all numerous times over each page of a website are soon coming to an end. And we can instead be focused on creating some great content, and display it in an extremely pleasing way for users. Not only is this more fun to create, but if done well, Google will reward marketers with better “relevance” and better exposure within the search positions. This will, in turn, force SEO’s to finally hang up their Black Hats for good.
Looking to the Future…
It is safe to assume that RankBrain is only the beginning, as computers utilizing machine learning, or A.I., will become smarter and more efficient at predicting semantic search, and SEO will evolve based on this.
So as long as RankBrain doesn’t one day decide to stop it’s focus on semantic search and revert instead on conquering the world, I think we can look forward to some cool advances in SEO – even in the near future.