AI has taken the tech world by storm, and now Google is using AI technology to enhance its search results for you.
What is Google RankBrain?
RankBrain is the name given by Google for an artificial intelligence system. RankBrain is an essential part of Google’s core algorithm and uses machine learning to help refine queries.
Google processes these queries with their RankBrain algorithm to deliver more precise and relevant results to the user.
How does it work?
The process begins when RankBrain collects the data from a variety of sources. This data goes through an interpretation model which then calculates and teaches itself over time, making the necessary changes to its algorithm.
Before RankBrain, Goole’s software engineers would code the algorithm by themselves. The algorithm would be tested and only then be updated to improve search rankings.
RankBrain, on the other hand, continually learns and develops its algorithm. Using a series of databases, it takes into account the location and the words of the query to deliver more precise results to the user.
In Google search, if Rankbrain does not recognize a word or phrase, the words are broken down into word vectors.
Search queries are then organized into mathematical entities called word vectors.
When RankBrain comes across an unknown query, it uses vectors to search for the words or phrases most similar and filters them accordingly.
Unknown queries Google would not have recognized before would be attempted to be understood by Rankbrain. Search results would be recorded over time and used to adapt and refine future search results to provide better user satisfaction.
In 2015 when RankBrain was implemented into Google’s search algorithm, it only worked with around 15% of Google queries.
Over time, however, with search results recorded and analyzed, RankBrain is present in almost all search queries.
Measuring search result satisfaction
RankBrain, while also recording search query results, can also measure user satisfaction based on user interaction.
The measurement of user satisfaction is the following:
Click-Through Rate
CTR measures the number of click-throughs a page gets from the search results.
Time-on-Site
This measures how long the user is on the website or calculates the average time a user spends on it.
Bounce Rate
Bounce Rate is the percentage of people who visit the website only to leave after viewing the initial page.
A bounce rate in the range of 26 to 40 percent would be considered satisfactory. 41 to 55 percent is roughly average, and a website which has a bounce rate of 70 percent and above is deemed to be unsatisfactory for users. However, these benchmarks can be completely relative to the niche and industry.
Pogo Sticking
Pogo Sticking is when a user visits the website only to leave and view another site.
Leaving most likely means that the user did not find the relevant information on the website based on the query searched.
Based on these SEO factors, RankBrain will change its algorithm by analyzing results with low user satisfaction.
From here, it will take the results with low user satisfaction and come up with a hypothesis on how to improve the situation in the future. If the results are positive, it will also change the algorithm accordingly, all by itself.
Impact on Digital Marketing
Branding
As RankBrain learns associations, it can benefit big companies as their brand becomes associated with certain products. Potential consumers making search queries on products will find themselves finding the same brands.
RankBrain also measures site engagement rate.
Even with an inferior link profile, more prominent brands are favourited due to being mentioned through different social media platforms.
Rank Brain and Mobile Phone voice searches
Voice searches occur when a user speaks into a device instead of typing keywords to search for queries. Search consultant Lorenzo Gutierrez tells us that by 2020, 30% of website sessions will be engaged by voice search as opposed to viewing sites through desktops and mobile phones.
RankBrain’s self-learning machine has made Mobile voice search results substantially more accurate over time.
Voice search queries are more conversational and typically longer compared to a typed query. It has been beneficial, therefore for mobile phone users as RankBrain evolves to understand these voice queries.
How RankBrain affects SEO
RankBrain has made SEO marketers concentrate on what website content it provides for Google’s search queries.
As keywords lose their importance, links and content have been the main drivers to boost rankings.
Pages which have only emphasized one keyword or keyword phrase have been suboptimal for rankings.
Content creators, therefore, are now targeting keyword variations and related keywords to improve SEO.
RankBrain has made content more natural as it needs to cater to search query results. Subsequently, content writers have been tailoring for people rather than algorithms. SEO marketers are creating content that sounds more human to match search queries which are conversational in nature.
In conclusion quality content that answers search engine queries as comprehensively and naturally as possible is paramount.
