Google Rankbrain - How it Works and How Influences Search Results in 2020

Up till 2017, Google was constantly changing their algorithms. These same algorithms that make your articles move up or down organic search results. According to research, Google’s engineers could introduce algorithm updates even 500 to 600 times per year. From time to time the company also released a bigger update such as Google Penguin or Panda that changed the rules of the search game for good.

Now you might be wondering: why Google’s algorithms had to be tweaked almost twice a day? Well, because they weren’t ideal. People wrote them, so it was likely that other people would find ways to go around them to make their business climb faster up the search results.

Today Google’s engineers don’t have to monitor their algorithms as closely as they used to have. This is all because of RankBrain and its ability to learn.

What is Google RankBrain

Google RankBrain is an algorithm that uses machine learning and artificial intelligence techniques (self-learning) in order to sort search results. It finds the most up-to-date and relevant pages that match your query and segregates them displaying relevant search engine results page (SERP). 

Its main purpose is to make you happy with pages that it returns. RankBrain does it by helping Google understand your query so that they can find what you are searching for.

Before the algorithm update was introduced, Google used other algorithms to display query results. But these other algorithms weren’t able to interpret the query using artificial intelligence.

Nowadays it is possible to incorporate your geolocation into search results and present you with pages that are personalized because RankBrain analyzes the information that it has about you and your previous queries and studies your query word by word to learn the true meaning behind your search. This smart approach to analyzing queries helps Google provide you with more accurate pages.

What’s more, the search engine can now easier and faster return accurate results to new searches that the engine “is seeing” for the very first time. This is possible because Google is able to associate a new search with your previous search queries.

How does the algorithm work?

This is how Google improved their algorithm’s thinking process. Below you can see the work that was being done at Google before the RankBrain was implemented:

And how it works now with RankBrain in charge:

How does Google RankBrain understand and interpret words that you’re using in your query?

A few years back, Google had a serious problem. Every day their search engine was processing searches that it never saw before. There were around 15% of these new searches, which is a lot, considering how many queries the engine processes each day.

Back then, every time someone searched for a phrase unknown to the search engine, Google had to guess what the user had in mind therefore the query results weren’t always accurate.

How the searching algorithm worked before RankBrain was implemented? Well, Google took the words of the user query and searched among the websites that used those words or some combination of them. Then it chose and displayed those pages that were the most relevant for the words used in the search.

The engine had a problem, however, when encountering new keywords or phrases. This was happening because Google had no knowledge of the user’s intent - the goal behind the search. As a result, for new queries, the algorithm had to guess and could return irrelevant pages.

So now you must be wondering, how does it work today? How does RankBrain know the meaning behind every user’s query, even the unique one? And how is it possible that it understands you like a human being asked a question?

All of this is possible because of machine learning algorithms. This is how they work: the self-learning algorithm matches known keywords with those that it sees for the very first time and makes a connection between them, it’s able to create associations. Based on word associations, the algorithm returns relevant search results.

If you are interested in learning more about how Google “learns” the meaning of new words, you can read more about this here.

How does RankBrain know whether you are satisfied with your SERP?

Google RankBrain not only understands your unique queries but is also able to test search results on users, tweak the algorithm’s parameters, and implement the changes to those parameters. All of this to serve you with more accurate pages.

RankBrain therefore constantly works on improving Google’s algorithms replacing engineers that used to go through similar processes before.

To assess the user’s satisfaction from returned SERPs, the algorithm checks for several indicators:

• Did the user return to the search engine results page to click on another link or tried searching again after briefly scrolling through your page? This indicator is called pogo-sticking and Google uses it to measure the relevance of your content to a particular query.
• How many people engaged only in reading one page of a given domain? How many of them explored the other domain pages after landing on one of them? The algorithm checks for the bounce rate.
• Organic click-through-rate checks for the number of people that clicked through a link from the search engine divided by the number of times the link has been displayed in SERP.
• The time that a user spent on a given website before returning to the browser page.

Because there are millions of searches per minute the search engine can customize its algorithm based on similar queries. If many people like a particular search result, the algorithm is going to elevate that page in the search engine ranking. On the other hand, if many people visit a page and are not satisfied with its content because they quickly go back to the SERP, RankBrain is going to pick it up as a signal of bad user experience with a given website. As a result, the page will drop in the search engine ranking.

Did the Algorithm change the way we do SEO?

My short answer to this question is - no. However, it depends really on your approach to content marketing and SEO before the implementation of Google RankBrain. There are different approaches to the search engine game. If you are an e-commerce owner you will use different SEO techniques than bloggers.

The only thing that you need to make sure you’re doing to optimize your website for Google RankBrain is to create high-quality content. But I guess I did not surprise you with this observation.

Google ranking signals - a key to a successful SEO strategy

You can use Google ranking signals such as pogo stacking, bounce rate, or organic click-through rate to increase the visibility of your page in Google because these are the factors that are taken into account by RankBrain when evaluating the relevance of your website. 

There are several ways for optimizing your page for Google RankBrain. In short, you need to make the users happy with the content on your site and to achieve user satisfaction you have to answer the following questions in regards to your website's content:

1. How to avoid a situation when a user bounces back to the search results right after visiting your page?
2. How to increase your organic click-through rate?
3. How to extend the average time a user engages with your website?
4. How to decrease the bounce rate and encourage a user to visit other pages within your domain?

Firstly, create useful and relevant content that will be helpful for your readers. You want to provide your audience with answers so they don’t have to look for information anywhere else. Make your content interesting and try to capture visitor’s attention from the very beginning to avoid bounce backs.

Secondly, optimize your meta titles and descriptions. They are crucial because they actually make users click through from the search engine results page and read your content.

Thirdly, make your content easy to digest. Add video content, infographics, and images as they have been proven to attract more attention than plain text and significantly prolong the average time a user engages with your website.

And last but not least, remember about internal linking. Linking to other sites within your domain helps decrease the bounce rate and extends the user engagement time as they are reading more of your content and browsing through new pages.

Different ranking signals can be used for different queries

Before RankBrain was implemented, Google considered different ranking factors when sorting and displaying search results. Some of those factors were: number and variety of backlinks, the length of the content, and whether it matched with the searched phrase, the exhaustion of the topic, and (on top of that) the keyword match.

This works differently now. Before the search returns results, Google checks what type of query has been passed to the search engine. Google’s algorithm has to first take into account the type of information that the user needs. If the user is searching for the latest news, the search engine can’t serve results that have the biggest number of backlinks because a new article didn’t have long enough to collect enough backlinks and a user is clearly looking for fresh content so the search cannot point the user to irrelevant, not up-to-date information. The user needs just recently published articles to appear at the top of the SERP.

Creating content on the Internet dominated by RankBrain goes down to concentrating on the user’s intent behind the search first. You have to focus on the user’s needs because the algorithm serves users first. And when you focus on that, without a doubt your search engine rankings will skyrocket.

Conclusion

RankBrain is the third most important Google ranking factor that has an effect on your success in Google search results. What is more, over time this algorithm is becoming more and more important. Therefore it’s crucial you optimize your pages for it.

Back to you! How are you going to optimize your website for Google RankBrain? What actions are you going to take today to make your content more user-friendly? Let me know in the comments below!