Google REVEALS How Rank Brain Works (2019 Hack Tips)
Since 2013 to date, 15% of the queries searched on Google are new and never seen before. That is six years in a row and still counting. Despite the fact that RankBrain was launched around 2015, about 4 years ago, this stat still remains the same. Why? We believe this is because of the rate at which search queries are increasing exponentially. How true are the speculations flying all around about RankBrain? Hang on!
WHAT IS RANKBRAIN?
RankBrain is a deep machine learning algorithm that works on queries that Google has never seen before. Basically, it endeavors to predict a better result for a new query that will satisfy the user’s intent. These predictions are based on similar queries from the past that are closely related to what is really being searched for.
Contrarily, RankBrain influence ranking, but it is not a ranking factor. It does not check the content on your site, nor take note of the number of backlinks you’ve got. It just focuses on how users interact with search results and the pages visited.
In fact, many sources at Google, including Gary Illyes, have confirmed that RankBrain is the 3rd most important algorithm used to rank and it comes immediately after Content and Links, respectively.
If RankBrain is this big and important, how does it work? Well, stay glued because we will get to that soon! But first, let’s take a look at what RankBrain do.
THREE MAJOR THINGS DONE BY RANKBRAIN
Emphatically, I must state again that, RankBrain is just one of the core algorithms used by Google to rank pages based on new queries searched by users. It acts on all queries, but doesn’t influence all results. RankBrain interpret users’ queries that do not have the exact words that were previously searched for. The 3 major functions of RankBrain are:
Understanding Search Queries
Ever since the Humming Bird algorithm update came to light on the 22nd of August 2013, users’ intentions have been the major factor considered by Google while processing queries and ranking search result pages. Previously, there were some search queries that Google had no idea on what it really meant. Consequently, wrong and irrelevant results were served to users that entered those queries. With RankBrain now, Google is able to fathom what result to serve for confusing searches, by comparing new queries with similar queries. It also helps Google to know which ranking signal/metric to prioritize for each query. Examples of such metrics are:
- Keyword matching
- Link diversity
- Anchor text
- Content freshness
- Domain Authority
- Engagement rate
- Content depth
- Related topics
Measuring User Satisfaction/Interaction With The Served Result
The process does not just end at simply understanding search queries, but it goes further instead. Brilliantly, RankBrain tries to detect if the user is satisfied with the given result by detecting user interaction with the results. Once some results are generated for every search query, users will definitely click on the given results to know if it can fully satisfy their intents. Once the results are not satisfactory, the users either move to the next site or simply enter a new search query. These actions send a signal to RankBrain that the results did not satisfy the intentions of the users. With this revelation, RankBrain will do away with those particular ones and generate newer ones once the same search query is entered again, and these new results are more likely to be correct.
Tweaking And Making Necessary Amendments To The SERP
RankBrain can de-rank pages that are not relevant to the search queries. In like manner, it also boosts the rank of pages that fully satisfied the intent of the users. Therefore, it is safe to say that
RankBrain is here to improve Google’s algorithm by constantly interacting with users and knowing the intentions behind every search query. This is possible because RankBrain is one of the core algorithms of Google that works on every search query given by users.
Now that the 3 major works of RankBrain have been discussed, the next question is ‘How does RankBrain pulls these off’?
HOW RANKBRAIN WORKS.
Experts suggest that, RankBrain only affects queries that have never been ranked before, as a result of this, everyone shouldn’t bother that much. But, if it is truly a deep machine learning algorithm that can tweak the SERP, is that statement still valid? No? The points below will answer that.
1.Understanding Search Queries
Google used to scan web pages to check if they contain the keywords being searched for before displaying any result. Most times, the results were usually flawed or irrelevant because this method of generating results was already abused by people that did what was termed ‘Keyword Stuffing’. To satisfy the users, Google needed to make some changes and stop ranking pages based on the number of keywords present on the page, but based on how relevant the content of the page is to the search query.
Now RankBrain, like a human would, always tries to understand the intent of the query. It does this by matching new keywords that Google has never seen before to keywords that Google has seen (i.e. similar search queries). On the surface, this may look simple but in the real sense, it is amazingly sophisticated and laborious at the micro-level. In simple terms, RankBrain as a refining algorithm uses machine learning to generate the best result to a query, even when it doesn’t know what the query means. It makes some guesses and I bet you, those guesses are accurately correct most of the time.
For instance, when a popular example used widely on the web, “What’s the title of the consumer at the highest level of a food chain”, was searched, with the help of RankBrain the results generated was correct.
Comparing this result to the one gotten when a much simpler query was searched (‘top level of the food chain’), the results were the same. Interestingly, RankBrain was able to detect the intent of the query, and successfully linked the long and vague query to the short and concise one, therefore generating relevant results.
The processes RankBrain follows to understand search queries can be summarized as follows:
• It pulls data from different databases to run the algorithm.
• Then, breaks down the queries into word vectors and gives each word an address using a mathematical formula (similar addresses for similar words).
• When an unknown query is processed by Google, the addresses given to the words are compared and used to generate a best fit to the query, returning many results that are related.
2. Measuring User Satisfaction/Interaction with the Served Results.
Google RankBrain does more than just understanding search queries, it probes further to determine if users are truly satisfied with the result for the search queries entered. If not, the results are discarded and new ones are generated for the next search. Using some signals, it is very easy for RankBrain to detect users’ interactions with the result pages and how well it engages the users.
Oftentimes, Google has made it known that hundreds of metrics are checked before ranking websites to be displayed on the SERP. In addition, these metrics are further broken into smaller units. The SEO table designed by SEL talks about the various metrics used for ranking by Google algorithms. But for RankBrain, these 4 signals are considered most of the time to measure user interaction from which user satisfaction is inferred.
• Organic Click-Through-Rate
When the results are displayed, the number of clicks that a particular page got is noted. Users commonly click on the first result, but if it happens, for instance, that the first 3 results were skipped and the 4th result got most of the clicks. This action will prompt RankBrain to boost the rank of the 4th result because it seems to be relevant. In short, the page that gets the highest number of clicks signifies how important the page seems to the users.
• Dwell Time
The length of time that the user spends on a particular page matters too. Nobody wastes time on things that does not interest them. Short dwell time does not always mean that the content is irrelevant, but it can happen because the content is too short and the user finished reading it in a short time. If the user is satisfied with the content gotten from the page, RankBrain will definitely notice it.
• Bounce Rate
On the contrary, if users leave a page almost immediately they come in, it means the users are not satisfied with the content on the page. So many things can put users off and cause them to leave the page, not just the irrelevancy of the content.
This term is used to refer to a situation where users keep checking the different results maybe because they are not satisfied with the previous ones they have checked. For instance, when a user searched for a query and the result was served, as usual, the first result is clicked on. But, if the first result is not satisfactory enough, the next result is checked, if it is still not satisfactory, he/she goes to the next. This process is repeated until a satisfactory result is gotten or a new query is searched.
RankBrain document and stores all of these interactions in it databases in order to serve relevant and better results when the same query is searched again. Once a satisfactory result is gotten based on users’ interactions, it is then incorporated into the core algorithm.
3. Tweaking And Making Necessary Amendments To The SERP
RankBrain is constantly learning and improving based on its frequent interactions with the users search queries and users’ interactions with the result pages. After numerous trials and errors, at a point, the users will be satisfied with the results generated from the search queries. Once, this is achieved, the changes are implemented into the algorithm and shows the same results for subsequent searches of the same query.
The ability of RankBrain to tweak the algorithm is one good reason why many believe that it is here to stay. This is very true, because over time, Google have expressed confidence in the results being generated by RankBrain and impressed with how well it is finally understanding search queries. No wonder it quickly rose to become the 3rd most important algorithm for ranking.
Since the inception of RankBrain, what has it achieved?
WHAT HAS RANKBRAIN ACHIEVED?
The major reason why RankBrain was launched was to improve the search experience and quality. With the ability of RankBrain to understand search queries better, Google is now able to serve relevant results to queries it couldn’t understand before. It is more like, when Google sees a query it has never seen before, it says “Hey Brainy, I don’t understand these words at all, what does it even mean?” RankBrain will respond with, “Hey dude, I am checking queries you have processed before that are quite similar to this one”. Once it is done, it then says, “Whew, I found this one, they are very similar. This is what the user wants to see”.
HOW HAS RANKBRAIN AFFECTED SEO?
RankBrain affects ranking, but has no influence with ranking. First, Google bots crawls and index a site without the interference of RankBrain. Initially, when RankBrain started, it only processed new queries that Google is unfamiliar with. But as time goes on, it got involved with every query that is being searched, new or existing. For new queries, it does the job of understanding the search queries and prompting Google on the best result to serve. But for existing queries, it actually sticks around to detect how users are interacting with the results served. Using the metrics discussed above, once it detects that users are not satisfied with any result on the page, it tweaks the algorithm and de-rank defaulting pages.
To be safe, once a site keeps up with the ranking metrics used by the core algorithm, you don’t have to be scared of what RankBrain will do.
PROVEN HACKS TO RANKBRAIN
RankBrain cannot be directly optimized since it does not even know what’s on the page. It only studies users’ interaction with the page to determine how satisfied they are. Therefore, satisfying the users’ queries will put your site in RankBrain’s good book. Consequently, RankBrain optimization still boils down to the same old, boring practices you hate to hear. Even with that, I will still mention them.
There are two major things Users interact with on a page:
Page design, and
Page design encompasses factors like site structure, ease of navigation, text color, background color, header tags, Ads placement, CTA buttons, etc. If apt attention is paid to these factors and they are done the right way, bounce rate will decrease to some extent. Also, the dwell time can increase to a reasonable degree. Although, page design is not the only factor that influences these actions and that is where ‘Quality Content’ comes in.
“Content is King”! This is a statement that is undoubtedly true. Even if you have an amazing page design and an excellent site speed, without content, they are naught. Only good content can get you relevant backlinks, increased traffic, returning visitors, and establishing your site as an authority in that niche. Without content, every other SEO efforts are going to fail in the long run.
An excellent content must contain the following:
Keywords: To be included in the Title tags, Alt tags of images, Headings, Meta description, Anchor texts. Moderation is very essential so as not to abuse this hack.
Deep and Valuable: To fully satisfy the readers, the content must be able to answer ALL the questions the readers have.
CTA: The call to action must be appealing as well before readers can be compelled to click/act on it.
Catchy Title: The title is the first thing that users come in contact with before the actual content. An attractive title will get the needed attention that will boost clicks.
Meta Description: Although, Google has said this is not part of the ranking factors, but still, the keywords are highlighted when the snippets are shown with the results. A good meta-description with highlighted keywords will definitely grab the attention of readers to increase the CTR.
RankBrain have not replaced the core algorithms of Google, instead it is only one component out of hundreds of various metrics that made up the algorithm. However, RankBrain is the brain behind users’ satisfaction which is a major factor considered when ranking is done. Also, because RankBrain is an AI that is continually learning, it may end up becoming the first metric considered for ranking.
Do you still have some doubts on Google RankBrain? Ask them in the comments below and I would answer every of your questions.