Google BERT Update – New Search System Introduced by Google

Google BERT Update – New Search System Introduced by Google


Google has rolled out a new search algorithm named BERT. This is the biggest change in Google’s search system since its last RankBrain update that was introduced almost five years ago. The internet is flooded with lots of rumors since the company confirmed the new search algorithm. The SEO community has different assumptions regarding this update however the changes in the search system will be seen shortly. BERT rolled out this week and soon the communities will see the impact of it on the search queries. Before discussing its influence on websites, let’s figure out what this update is.

What is Google BERT?



BERT, that stands for Bidirectional Encoder Representations from Transformers, is the latest update rolled out by Google. It is an open-source neural network-based bidirectional and pre-trained deep learning algorithm that relates to Natural Language Processing (NLP). The algorithm is presently targeting English queries only and will be updated for other languages with time. The concept of BERT can be considered as a language model that was prefaced by Google in 2018 in a paper that was published by the researchers at the Google Artificial Intelligence Language. It is not targeting the search queries but also the featured snippets.

As per Google, BERT provides state-of-the-art results by understanding the distinctions and context of the words in searches. BERT understands the intent behind the search queries and delivers the results accordingly. It helps Google to understand queries that are conversational and more human-like. Let’s understand this concept with an example:

Suppose you searched for a query, “I need to travel to USA for FIFA cup”. In this particular query, the word ‘to’ and ‘for’ and their relation with other words are also important. In the previous approach, Google used to ignore such words and delivered results related to ‘FIFA world cup’ or’ USA’s performance in FIFA’. As you can see in the below image,


After the update of BERT, Google understands the full context of words and provides more accurate and related results. It is good for longer and conversational queries, where the prepositions are significant in the search. It will enhance the user experience by letting people search in natural way and delivering relevant results.

What will be the impact of BERT?


BERT will impact on 10% of the search queries. This simply refers to one in ten search queries. BERT is designed to focus on the user’s intent of the search and thus Google is going to look for the content that is relevant to the user’s search query.

Can SEOs optimize for BERT?


SEOs are looking for ways to optimize their sites according to the BERT algorithm. According to Google, BERT is a lot like RankBrain in terms of the machine learning algorithm. However, it is not the replacement or updation of RankBrain; it will be used in conjunction with all other algorithms. 

SEOs can’t really optimize their websites for BERT. There is some misinformation that websites can rank better by targeting long-tail keywords in their content. But this is not the complete reality. All they have to do is to write content for their users. The algorithm is for Google to get a better understanding of the search queries and delivering relevant results.

In a nutshell


Google is taking many steps that can help the user to use the search engine in the easiest way. But it may bring more competition in terms of ranking. To avoid the negative impact of any Google algorithm, the websites should focus on writing the content from users’ perspective. It cannot be said that BERT will not affect ranking but there is no specific way to optimize sites for BERT except that you should have your user in mind.

Comments