There was a lot of talk last week in the Search Engine Optimization (SEO) community about an article in Bloomberg News on Google’s use of artificial intelligence to return search results for a very large fraction of the millions of search queries it processes every day.
Greg Corrado, a senior research scientist with Google, officially introduced Google’s artificial intelligence (AI) system nicknamed “RankBrain”. Google has been a big promoter, and one of the largest corporate sponsors, of AI—and has invested heavily in it for videos, speech, and translation. According to the Bloomberg News article, “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.”
RankBrain apparently helps the search engine deal with 15% of queries a day; for now, these are limited to queries that its systems have never seen before. The article gave an example how it is “adept at dealing with ambiguous queries”. This is in line with the last major algorithm update (dubbed Hummingbird) which had a significant effect on semantic search. The article also stated that in tests, the AI system regularly beat the company’s experts at page selection—RankBrain had an 80 percent success rate, vs. the Google search engineers with a 70 percent success rate.
In the article, Corrado went on to explain how RankBrain is one of the “hundreds” of signals that go into an algorithm to determine what results appear on a Google search page, and where they are ranked. He also said that in the few months since it has been deployed, RankBrain has become the third-most important signal contributing to the result of a search query.
Moreover, Google recently revealed in its third-quarter earnings report that mobile has, for the first time ever, surpassed desktop search.
Many of those mobile searches likely come through speech (there’s no available data on this), which also likely includes a higher percentage of unrecognizable queries compared to typed queries. As AI is being used in speech, this will likely result in higher quality results for mobile users. This could give Google an edge over Apple (Apple uses Bing as the default search engine for Siri). And with RankBrain being such a huge success so far, this likely means that over time, the AI system will start processing even more queries, and we should start to see even more quality search results in Google across both mobile and desktop.
Although search results being generated by AI is cool and futuristic, when it comes to search engine optimization, the old reliable methods of SEO hold true; ensuring the basic SEO best practices are followed is still as important as ever (these can be found in Google’s Search Engine Optimization Starter Guide from 2010). Keeping up-to-date with current developments in SEO including schema markup is also recommended—this involves following official Google blogs (Google Webmaster Central Blog and Google Inside Search) and other industry blogs such as Search Engine Land and Moz.
The diagram at the start of this post is from a 2013 Google Inside Search blog post that illustrates how search has evolved since Google’s inception in 1997. The title of that blog post is ‘Fifteen years on—and we’re just getting started’. Looks like Google’s keeping true to their word!
* Update: After this post was published, Google published a post titled ‘The Google app now understands you a little better—complex questions welcome‘ which covers advancements in query recognition through spoken search queries. Check that out if you are interested in this subject!