Meet RankBrain, the New AI Behind Google's Search Results
As we all know, Google is constantly looking to provide more relevant results for its users -- hence, the regular algorithm updates that frequently frustrate webmasters and anyone else's SEO efforts.
The Pagerank algorithm that founders Sergey Brin and Larry Page introduced in the early days of Google was a step in the right direction, but it certainly wasn’t the ultimate solution for improving the quality of search results.
In fact, the search giant recently unveiled a new AI (yes, that stands for artificial intelligence) called RankBrain to help the engine better understand the queries users type into the search field. The real intention of this AI wasn't to change visitors' search engine results pages (SERPs) -- rather, it was to predict them.
As a machine-learning system, RankBrain actually teaches itself how to do something instead of needing a human to program it. In other words, RankBrain will help Google provide valuable results for search queries that it hasn’t encountered before.
New search queries account for about 15 percent of all the queries that Google receives, which poses a significant challenge to the engine to provide relevant results, compared to known queries. Now, RankBrain is here to help solve this problem; so brands that rely on SEO would do well to understand the impact of this new AI.
How RankBrain will change search results
RankBrain is not a new algorithm, but rather one of the hundreds of signals that make up an algorithm. It analyzes web pages related to every possible search query and has the ability to make connections between and among different words or phrases, so that the algorithms used can provide the most relevant search results.
For example, the algorithms used might examine content related to iPads and then draw a connection between the terms “Apple,” “iPhone,” “iOS” and “Steve Jobs.” Previously, Google would look only at pages with the word “iPad,” but RankBrain understands that content containing related words will be highly relevant to the user’s search queries.
RankBrain is also useful for people who are searching for uncommon or ambiguous words and phrases, as the engine can deduce what those people are looking for, based on their chosen words' contextual relevancy to different search terms.
According to Jack Clark, writing for Bloomberg on the topic: “[Rankbrain] uses artificial intelligence to embed vast amounts of written language into mathematical entities -- called vectors -- that the computer can understand. If RankBrain sees a word or phrase it isn’t familiar with, the machine can make a guess as to what words or phrases might have a similar meaning and filter the result accordingly, making it more effective at handling never-before-seen search queries.”
How effective is the new AI?
RankBrain was unveiled only a month ago, but the results so far have been very impressive. Greg Corrado, a senior scientist at Google who helped develop the AI, said that the new algorithms have already become the third most important ranking factor and that they’re significantly improving the quality of the search results:
“The other signals, they’re all based on discoveries and insights that people in information retrieval have had, but there’s no learning,” said Corrado. He continued: “I was surprised . . . I would describe this as having gone better than we would have expected.”
How will RankBrain evolve?
It’s clear that Google intends to introduce new updates to the AI in the future. The developers behind RankBrain believe that it has a lot of potential, but they also understand the importance of testing it first to ensure that it delivers the quality results they’re hoping for. Ultimately, RankBrain needs to be able to accurately draw connections and forecast future searches before it can be expanded and improved upon.
The program so far has proven that it’s exceptionally adept at doing just that. In one instance, Google conducted an experiment to see if RankBrain could make accurate predictions of which pages would rank highest as compared with a team of search engineers. The human searchers were correct 70 percent of the time, while RankBrain boasted an impressive 80 percent accuracy. And while it may be premature to say that this new algorithm is already superior to human users, it’s clearly a promising piece of technology.
If these results continue to be consistent, it’s likely that Google will invest even more of its efforts into using RankBrain, potentially giving it more and more control over search rankings.
Should brands relying on SEO be nervous?
Anyone who has worked in SEO for a few years understands that the changes at Google have had -- and will continue to have -- notable effects on the search engine rankings. The Panda and Penguin updates, for instance, caused significant penalties for a number of sites, and some were completely de-indexed overnight. Certainly, fears about any new changes to the engine’s algorithms are warranted.
But while the RankBrain algorithm has received a tremendous amount of press already and has already made some changes to the search engine rankings, its impact hasn’t been consistent on all brands.
The brands most likely to notice a difference in their rankings are those trying to compete for keyword phrases that Google hasn’t encountered before. For example, a company offering a new product or service that few people have searched for may see a ranking shift. Additionally, some brands might see a ranking shift for their own names as well -- particularly if their names include some commonly used words.
It’s worth noting that future updates to the RankBrain algorithm could have a more significant impact, so keep a close eye on the new AI changes that Google is rolling out, as they could drastically alter the future of search.
Are you concerned about RankBrain’s potential impact on your business’s search rankings? Or do you think the whole issue is being blown out of proportion? Share your opinion by leaving a comment below.