Have you ever wondered how Google is able to return search results? The exact science behind how search works has always been a mystery, and today the mystery seems straight out of a sci-fi novel. Enter RankBrain, Google’s artificial intelligence (AI) technology, which aims to improve search in a vastly new way.
Let’s explore how it might change your world, and let’s start with the simplest question: What is RankBrain?
RankBrain is a machine learning technology developed by Google.
Announced in October of 2015, the RankBrain system was developed to help improve search results by making them more relevant to what searchers are looking for. The company has always touted the mission of “user first.” RankBrain aims to use advanced technology to assist its engineers and scientists in improving the search algorithm for its users. The days of Google not understanding what you’re looking for are (possibly) nearly over!
This new development marks an exciting time in history when machine learning and artificial intelligence are brought to center stage. Straight from science fiction movies and novels, the concept of machine learning and AI is somewhat new to most the American public. (But not all of it. Read about pioneer Ray Kurzweil.)
Experts are divided in how (or whether) RankBrain will affect rankings.
RankBrain is a form of machine learning and artificial intelligence.
What is machine learning and artificial intelligence? First, machine learning is a concept you’ve probably heard about. It’s when computers learn from their behavior and teach themselves how to do things instead of being taught by humans or following detailed programming.
Machine learning is used alongside AI. AI refers to a computer program which is designed to learn and make new connections based on what it teaches itself. The terms are almost synonymous with each other, so you may hear them used interchangeably.
RankBrain does not replace Hummingbird, Google’s current search algorithm.
RankBrain shouldn’t be confused with a replacement for the existing Google search algorithm in place. Google announced in the middle of 2013 that Hummingbird was the “fresh and new” engine responsible for ranking pages. Hummingbird was developed to focus on word meaning and take into consideration the entire query and its context, together with semantic clues and proof words, not just individual words. RankBrain aims to improve on this technology already in place, by introducing machine learning into the mix.
Imagine an engine that not only understands the meaning of your words and long tail phrases, but also is able to predict what you really want to see – even if you’re not using the exact right words.
RankBrain is one of over 200 signals Google uses to rank web pages.
It’s true, RankBrain is just one of MANY signals Google uses to rank pages. Google has always been hush-hush about what those exact 200 signals are and how important each is, but we can speculate that RankBrain is extremely important. That importance will only grow as tests and research continue to show positive results.
It is the third most important signal for ranking.
In fact, RankBrain is the third most important signal for ranking as reported by Greg Corrado, Google Senior Research Scientist. While they won’t tell us what the top two are of course, he did state in an interview with Bloomberg that RankBrain is important enough to make the top three. (Editor’s note: In a recent Q&A with Search Engine Land, Andrey Lipattsev, a Search Quality Senior Strategist at Google, said the other two factors were links and content.)
Testing is still underway as to exactly how important RankBrain is to improving rankings, but only time will tell. For now it’s just one piece to the ranking puzzle.
RankBrain is already starting to improve search results.
Fast Company reported that Eric Enge and his team at Stone Temple evaluated over 1.4 million Google search queries from July of 2015, before RankBrain. They looked into the subset for which Google didn’t provide appropriate results.They found that of the queries that Google gave poor results to pre-RankBrain, when submitted again to Google with RankBrain, 54.6% were answered correctly. .
The true results of RankBrain will take some time to see, but it’s encouraging that expert SEOs like Enge are already starting to feel its power.
AI (like RankBrain at Google) is here to stay.
Bloomberg reports that extensive dollars have been funneled into research and development projects at Google to unlock the power of artificial intelligence. “Things are going to get weird over at Google,” said Greg Corrado. It sounds like AI at Google is making its way onto center stage, but time will tell. Google has a way of changing its mind – often – when it comes to search algorithms.
What do you think about RankBrain and the future of search? Share your thoughts with us in the comments below.
SEO is necessary, complex, and confusing. SEO helps expand awareness of your brand, attract new prospects, grow your customer base, create new revenue channels, and grow existing ones – however, many marketers feel left in the dark when it comes to SEO. Download our eBook, “Your Most Common SEO Questions, Answered,” to learn the most common questions about SEO from marketers and business decision-makers.