Announced in late-2015, RankBrain is Google’s latest big algorithm when it comes to search and SEO.
And as a marketer, understanding how Rank Brain works is key to rank at the top.
Don’t believe me? Take it straight from Google!
In a interview with Bloomberg, Google expressed that RankBrain is the “third most important factor” when it comes to determining the order of the results they show. (The other two are content and backlinks).
In this article I want to dive into what RankBrain is, understand how it operates and propose a strategy to skyrocket your rankings.
Ready to beat Google at its own game?
- 1 What is RankBrain?
- 2 Let’s dive deeper into how RankBrain works
- 3 How does RankBrain measure user satisfaction?
- 4 How does Google Rank search results?
- 5 How to target keywords that RankBrain will love
- 6 To beat RankBrain, optimize for medium-tail keywords
- 7 Conclusion
What is RankBrain?
Rank Brain is an algorithm launched by Google that helps it sort its result pages using machine learning.
(Machine learning is a way for computers to learn and extract information based on data that they encounter, as opposed to coders explicitly declaring that knowledge.)
Previously, Google Search Results were tweaked manually by engineers based on metrics they thought were useful (i.e. CTR, dwell-time or load speed).
Now, with RankBrain, Google is letting an AI tweak its results autonomously, on the fly.
For instance, for a particular keyword like “chocolate chip cookies”, RankBrain may decide that backlinks are not as important as reviews, so it can modify the importance that those metrics carry on the SERPs. (Search Engine Result Pages)
But how does RankBrain know which metrics are relevant for a particular keyword?
It doesn’t. Rankbrain learns by means of experiments.
Following the cookies example, RankBrain adjusts various SEO metrics until it finds a formula that leads to the highest customer satisfaction (which could be lowest bounce rate, for instance).
Through Google’s own experiments, they found that RankBrain was 10% more accurate than Google’s own engineers when it came to predicting the best page to show.
RankBrain is now deployed worldwide in all the millions of search queries that Google receives each second.
RankBrain is a part of the much larger, Google Hummingbird algorithm.
According to one of the heads at the Google Machine Learning division, Jeff Dean, RankBrain affects the actual rankings “probably not in every query but in a lot of queries.”
Let’s dive deeper into how RankBrain works
As I mentioned before, RankBrain constantly organizes the SERPs to find the most accurate rankings for a particular search term.
RankBrain uses semantic analysis to understand what your query is all about.
This means that RankBrain no longer just looks at keywords (and keyword density) but rather, it tries to understand the meaning behind your search.
For instance if I ask, “first product by apple”, then RankBrain produces the following accurate results.
In the past, pages where the keywords “first”, “product” and “apple” appeared the most and had the most relevance would be shown.
Same deal if I search for “who sings Gucci Gang.”
As evidenced, Google cares more about the context and meaning of your search.
To do this, RankBrain groups words into concepts and finds pages that covers those concepts in depth.
It also takes into account things like user location. For instance, if you search for the “World Cup 2018 location” and are located in Russia (World Cup host) , then it might show map directions. If you are located in the US, then it might just show information about the city where it is being hosted.
(Keep reading to see how this affects SEO keyword targeting in a RankBrain-first world)
How does RankBrain measure user satisfaction?
Google’s ultimate goal is to show you the best possible set of pages and user satisfaction is at the core of Google search.
Although the actual satisfaction metrics have not been released officially by Google, we can make assumptions as to what those are.
If I had to guess, then I would say RankBrain looks at:
- Organic CTR (Click Through Rate)
- Time-on-Site (aka. Dwell time)
- Bounce Rate
- Domain Authority
- Pogo Sticking (when you quickly leave a page and go back to the SERPs)
Based on these SEO factors, RankBrain constantly shifts pages around until each page has achieved its deserved spot on the SERPs.
For instance, let’s say that most people
- click on result #1,
- skip results #2 and #3,
- and click on result #4, spending a lot of time in the #1 and #4 results.
RankBrain notices this and gives result #4 a boost next time someone searches for that keyword. It also lowers results #2 and #3 because they were not appealing.
With the billions and billions of keyword searches that Google receives, RankBrain has a lot of data to experiment and pick definitive winners.
How does Google Rank search results?
As explained before, the results are ranked by 1) keyword relevancy, 2) number of backlinks and 3) Rankbrain. In this article I go more in-depth about how you can reach the top of the SERPs.
How to target keywords that RankBrain will love
It seems like the days of long-tail keyword targeting are gone.
Back in the day, it made sense to create content for different but closely related long tail keywords like:
- Best credit cards for students
- Best student credit cards
And have each page meta tags optimized specifically for each long tail variation.
Nowadays, that SEO technique is dead.
Because with RankBrain conceptual search, long tail keywords are now grouped into concepts rather than specific wordings. The previous example looks like
(Best, Top) <–> (Student, College) <–> (credit cards)
And any possible combination of these keywords leads to practically identical search results.
Therefore, optimizing for long-tail keywords is not effective in 2018 anymore.
So what’s the alternative?
To beat RankBrain, optimize for medium-tail keywords
Unlike long-tail keywords were search volume and competition are rather small, medium-tail keywords do generate a good amount of traffic (and thus healthy competition).
They are at the sweet spot between almost-impossible, broad keywords like “SEO” and too narrow like “how to do your own seo for free.”
A medium-tail alternative for the aforementioned example would be “how to do seo”.
Although that post won’t rank very high for just “SEO”, it will rank high for tons of long-tail variations. Provided that blog post is written perfectly, that is.
Here are other articles that will interest you it comes to doing search engine optimization with RankBrain:
- The best SEO tips ever. Period.
- Top On-Page SEO factors that RankBrain doesn’t want you to know
- The SEO Tools you need to use if you are serious about rankings
- How to find keywords for Rankbrain
Bonus: How to make your medium-tail keywords even better
To better help RankBrain understand what your blog post is all about you should include natural variations of your keywords throughout your text (known as LSI keywords).
For example if you are writing about “ab exercises”, you can mention keywords like “ab crunches”, “core exercises”, “abdominal workout”.
All these LSI keywords help RankBrain associate the concepts with ease.
Here are 11 other on-page SEO factors you should consider if you are serious about organic traffic.
RankBrain is a powerful algorithm that constantly tweaks itself to provide you with the best results based on user intent.
SEO in 2018 requires knowledge of RankBrain. Keyword targeting is no longer a long-tail game but rather a much deeper contextual battle.
If you want to rank at the top, you need to do the following:
- Write great, in-depth content
- Include variations of your keywords
- Provide such amazing value to your readers so that they stay long on your site and keep coming back for more
If you liked this post, be sure to check out my SEO archives.
Also drop a comment if you liked this post. I did a ton of research and it makes my day to know that someone appreciates it 🙂