Approaching RankBrain Intelligently

If artificial intelligence seems like a pipe dream, consider this: you’ve (probably) already been using it for months. In fact, your last brush with AI is as recent as your last Google search. A machine-learning form of artificial intelligence, known as RankBrain, is currently Google’s third most important ranking factor for webpages in search results. What does that mean for those in the SEO business? As always, it’s survival of the fittest, and only those who adapt to the new search landscape will succeed.

 

Search Gets Smarter

In October 2015, Google made an algorithm announcement that shook the SEO world. Unbeknownst to anyone outside of Google, RankBrain had been baked into the larger Hummingbird search algorithm over the previous months. What’s more, out of over 200 ranking signals, RankBrain was quickly deemed one of the factors that matter most. Google typically doesn’t announce algorithm changes unless they’re confident the update will be successful. The news signaled a big step towards a future that relies less on human judgment, and more on self-learning machines.

 

Photo Courtesy: John Trainor | https://www.flickr.com/photos/trainor/451799414
Photo Courtesy: John Trainor | https://www.flickr.com/photos/trainor/451799414

In order to understand the significance of RankBrain, you first need to understand the basics of artificial intelligence. What sets AI apart from other computer systems is its ability to make connections and learn (aka self teach) without human intervention. For example, if you show an AI machine thousands of pictures of trees, it will eventually understand the concept of “tree” and will be able to identify something as a “tree” without any assistance. For Google’s purposes, this meant months, if not years, analyzing millions of queries and connecting them with relevant results – Easier said than done!

RankBrain is modeled on an actual human brain, and because of this, Google can never be 100% sure why the technology performs a certain action. Paul Haahr, one of RankBrain’s head engineers, has stated that they understand how RankBrain works, but not exactly what it’s doing. Just like a biological neural network, individual connections – in this case, everything from phrases to images – are strengthened or weakened over time. This factor comes in handy for sorting complex, multiword search phrases that can be interpreted several ways.

 

A Better User Experience

If you try a Google search with the query, “What is the weight of Trump’s victories?” you get the following results:
rankbrain results

Notice that Google was able to understand that the word “weight” in this context means “importance,” rather than a type measurement. The highly ranked results are all relevant to the query, even though the phrase takes a more human-like thought process to interpret. What this means is that RankBrain can essentially translate search phrases, as well as webpage content, with a degree of accuracy that takes context into account.

Interestingly, RankBrain assigns a relevance score between 1 and 10 to every page in its database. For those of you familiar with PPC, this probably sounds familiar. It works in a similar way to the quality score metric in Google AdWords. The most relevant organic search results are rewarded with high ratings, and in turn, high rankings. It’s part of a larger trend to create a better overall search experience – Google’s recent search quality guidelines made it clear that websites that provide specific answers to specific questions will be rewarded.

While RankBrain ignores traditional ranking factors, such as backlinks, it does factor in click-through rate. As with a quality score, Google calculates the expected click-through rate for a particular search result. If your webpage gets more clicks than expected, it is rewarded. Underperformers are penalized with lower rankings.

Image Courtesy: WordStream | http://www.wordstream.com/blog/ws/2016/03/16/rankbrain-seo
Image Courtesy: WordStream | http://www.wordstream.com/blog/ws/2016/03/16/rankbrain-seo

 

Making Use of RankBrain

The mechanisms behind RankBrain are fascinating, but you may be wondering how your business or SEO agency can take advantage of this newfangled ranking signal. To succeed with RankBrain, your website MUST create content that fills specific needs. This could be as specific as the answer to a common question. For example, a car repair shop website could benefit from a “What’s This Noise?” section to answer common questions they receive from customers with older cars. Starting with extremely relevant content naturally brings more clicks to the website, which RankBrain rewards again with a higher relevance score. In this sense, you’re better off writing more comprehensive content on topics you know the most about, rather than branching off into niches and providing less satisfying answers to searchers.

RankBrain’s emphasis on click-through rates means that exceptional title tags and meta descriptions are more important than ever. Anyone who’s ever written ad text for AdWords knows there’s a fine balance between including keywords and writing sentences that engage people. The best way to get users to click on something is to get them to feel something, and copy that resonates emotionally is much more powerful than simple descriptions or keyword-stuffed headlines. If you can include both strategic and provocative elements in your title tags and meta descriptions, you will probably see a healthy increase in your click-through-rate.

 

Does It Work?

Because RankBrain has been active for months, the big question is whether it actually works. A recent study analyzed complex Google searches before and after the algorithm change took place. It found that Google was able to improve the quality of search results 54.6% of the time with RankBrain. Additionally, Google was more likely to add a map and improved answer box with the assistance of its AI algorithm update. This study shows that if people are looking for your website information, they’re much more likely to find it – even with an unclear search query. For content creators, this means what they produce is much more important than how it’s worded.

 

The Future Starts Now

There’s no way to ignore RankBrain – it’s an integral part of search and will continue to gain importance. Google recently hired John Giannandrea (5), the company’s former head of artificial intelligence and machine learning, as its new head of search. The move suggests that Google is ready to give AI more responsibilities when it comes to organizing search results. For anyone in the SEO industry, that means it’s important to pay attention to RankBrain’s progress and implementation. You may not have started your career thinking you’d be an AI expert, but it could be where the industry is headed.