In September 2019 the Californian giant officially launched its new algorithm, considered one of the most important ever developed since its beginning in 1998, called BERT – Bidirectional Encoder Representations from Transformers.
According to Google itself, the impact of this update on its ecosystem will only be less than RankBrain, which was the algorithm based on machine learning in 2015. We will talk about it below …
What changes with the BERT algorithm?
Each year Google makes hundreds of changes to its search algorithm. In 2018 alone, Google reported 3,234 updates, that is, an average of almost 9 per day and more than 8 times the number of updates in 2009.
It turns out that in the midst of these small changes, Google occasionally releases an important (major) algorithmic update that significantly affects search results, as is the case with Panda in 2011 and Penguin in 2012. We will also talk more about them soon below.
In the case of BERT, to better understand what this change represents and how much it impacts on the lives of consumers and brands, let’s go back in time, back in the era of the iPhone 3G, to understand how Google has been changing over time.
Most important updates to the Google algorithm in recent years.
Panda 2011
It was the update that marked the beginning of SEO and Content Marketing, since the algorithm started to work relating the quality of the content.
This update brought a better search and delivery of more refined information in the search results. With Panda, Google also introduced a quality rating for pages.
Penguin 2012
To further improve the quality of the content delivered to the user, this update focused on eliminating spam links and some SEO practices that circumvented previous robots.
Until 2012, Google paid close attention to the volume of visits to a page to rank it. That is, pages with low content quality, but with a high volume of visits due to SEO techniques ended up having a lot of relevance. Penguin helped to change this scenario a little, making it difficult to accept unacceptable SEO techniques.
Hummingbird 2013
In 2013, the semantic research format entered the scene. This means that, as of 2013, Google has started to perform its searches according to the user’s profile and history, offering more variables.
At that time, Hummingbird (Hummingbird) started using semantic search to provide results that corresponded to the researcher’s intention, helping Google to understand conversational language, such as longer queries. (Long-tail keywords)
Pigeon 2014
In 2014, it was the turn of local businesses. With Pigeon, Google updated the local ranking signals to provide results “close to me” for users.
Based on the GPS information of the devices, the algorithm improved the distance and location classification parameters.
Mobilegeddon 2015
In this update, Google increased the ranking of mobile-friendly pages in mobile search results. Users are more easily able to find relevant, high-quality results, where the text is readable without touching or zooming, the touch targets are spaced properly and the page avoids non-reproducible content or horizontal scrolling.
RankBrain 2015
Also launched in 2015, RankBrain was an algorithm that introduced the search engine based on Machine Learning.
Fred 2017
This was a more comprehensive update and focused on quality on several factors. Always seeking to improve the quality of the content found, with Fred, Google started to filter even more sites that used SEO techniques for commercial purposes.
What about BERT in 2019?
BERT is considered a mechanism for responding to queries that are not possible to anticipate.
According to Pandy Nayak, the goal of BERT is to decipher what the user wants to look for on the web, no matter how he types or combines the words on the keyboard ”.
BERT is also designed to better understand the language in which we speak every day. In essence, Google search is based on understanding the language.
Results page changes
For a long time, Google’s search results used to be simple: search results in the middle of the page, ads on the right.
However, a few years ago, Google tried to place some ads above the search results, and they worked very well for advertisers.
Based on these analyzes, the results page changes with the arrival of BERT.
As Google said, “Our analysis shows that ads above the search results are clicked 14 times more often than those on the right, so it makes sense to remove those ads, but add one more at the top of the page. It appears only in the results that we consider “the most commercial”, that is, we are confident that users are trying to buy something. ”
For the user, this means that fewer total ads are shown on the search results page, as the amount paid by Google will be paid per click and not for views.
As for the advertiser, there is a reduction of ads on the results page, which increases the cost of ads at the top and greater competition in the Google ad auction process. It will also force the best positioned to move to other paid ad locations.
In addition, most commercial searches now display at most one or two “natural” search results without scrolling. Google says that these changes will improve relevance to consumers, but that remains to be determined.
What does this mean for the entire digital ecosystem?
To my understanding it simply means that if you want to “rank” on Google, your title has to be top, your fantastic content and your highly planned and monitored Ads strategy more and more.
As I said above, BERT is an open source neural network that has been trained to process natural language. That is, Google learns from you and you learn from it.
For the user, it increases the confidence that the information searched is relevant and quality.
For Marketers and SEO specialists, in addition to producing a lot of Links. With the new Google BERT, the generated content must be optimized for people, not Google.
In addition to creating memorable content for the user, we must not forget more and more Google also gives importance to Branding and user experience.
Therefore, creating a website with thought flow of information, which optimizes your customer’s purchase funnel is key to your online performance.