Introducing Google RankBrain

Here come the machines! Are you ready for artificial intelligence (AI) in the world around you? Put your seatbelt on and get ready.

Companies that have purchased AI firms include Apple, Yahoo, Intel, Dropbox, LinkedIn, Pinterest and Twitter. Toyota is investing $1 billion in an artificial intelligence project in California. IBM has been investing in Watson for years. It’s not just companies making the investments. Influential tech leaders such as Elon Musk and Mark Zuckerberg are investing their own money in AI, as well. It feels as if everyone wants in on the action.

According to Quid, an AI company that accelerates research and insights through its analytics platform, AI has attracted more than $17 billion in investments over the past five years. In fact, Quid predicts investments in AI to grow by 50% in 2015 over last year’s levels.

Google’s Obsession

One company that is talking non-stop about AI and machine learning, and if anything, appears to be obsessed with the potential, is Google. The search giant spent $500 million to acquire the AI startup DeepMind last year. Overall, Google has invested in AI in the areas of videos, speech, translation and, now, its core business of search.

Google CEO Sundar Pichai recently stated that advances in machine-learning technology would soon have an impact on all Google products and services—across the board. Pichai underscored the importance of this new focus, stating, “We are rethinking everything we are doing.”

Introducing RankBrain

Greg Corrado, a senior research scientist at Google, revealed to Bloomberg that a large fraction of Google searches performed over the past few months have been interpreted by Google’s own machine-learning artificial intelligence platform, nicknamed RankBrain. The AI technology enables Google to interpret the 15% or so of searches that the algorithm has never encountered before.

RankBrain uses artificial intelligence to convert written language into mathematical entities called “vectors,” which a computer can understand. In this way, when the algorithm encounters a first-time query, it can make associations with words or phrases that most likely have a similar meaning, and then filter the search results accordingly.

RankBrain is already proving to be effective in making search results more relevant, and in fact, is exceeding the expectations of Google’s own search engineers (the ones who craft the algorithms that underlie search as we know it). In one test, the engineers were asked to identify pages they thought that Google’s algorithm would rank at the top. The engineers were correct 70% of the time, while RankBrain had an 80% success rate.

What RankBrain Means for Your Business

How much of an effect is RankBrain having on Google search? Corrado says that RankBrain is now the third most important signal contributing to the result of a search.

Google has spoken through the years of having more than 200 ranking signals, and that each of these may have up to 10,000 variations. The AI platform is not replacing Hummingbird, but it looks to now be an important signal that Google will factor in evaluating the relevancy of content when ranking query results. Although there are clearly many, many factors for a single search ranking, considering that RankBrain is already the third most important signal is actually astounding.

Do you need to change your SEO strategy as a result of RankBrain? No, not at this time. RankBrain is simply helping to make search more accurate.

Given Google’s obsession with AI, though, it behooves your company to keep an eye on future developments, as AI is only going to become more and more integral to online search. The machines are coming, and your business needs to be prepared.

Photo Credit: Samuel Zeller