Better understanding long-tail search queries and presenting more relevant results – that is the goal of BERT. For five years, it has been the biggest algorithm update and affects one in ten search queries. So far, however, it only applies in the USA. In Germany, only the featured snippets have been affected to date: When BERT will fully arrive is therefore still open. But what exactly is BERT? Why is the update coming? And will anything change for SEO optimization?

Google Update BERT: Better Understanding Search Queries

What is BERT?

BERT stands for „Bidirectional Encoder Representations from Transformers“ and aims to do one thing: better understand human language. BERT improves its language comprehension through a self-learning neural network („Transformers“). But how does BERT learn? By considering words in context; that is, what comes before and after the search term („Bidirectional“). Nevertheless, BERT does not comprehend language like a human, but merely simulates it for its purposes. Consequently, language is only represented („Representations“) on the computer.

But what exactly does this mean for SEO optimization? Google wants to better understand the search query by treating words as interconnected – and no longer word by word. An example: „Can you buy bread rolls at the gas station?“ Before BERT, Google would search for bread rolls and gas station; that is, locally for a bakery and the nearest gas station. With BERT, the entire question is now in focus, and with it, the appropriate answer. Consequently, the focus is no longer just on keywords, but also on small words – prepositions, adjectives, negations, etc.

Why is BERT coming?

But what makes BERT so important for Google? The reason lies in the search queries themselves: They increasingly resemble spoken language. Search queries are becoming more human and are increasingly approaching normal communication. This is made possible by the development towards voice search – in two years, up to 50 percent of queries are expected to be verbal. But user attitudes are also changing: They want lightning-fast answers to complex questions. This is only possible with long-tail search queries – they already account for 70 percent of search queries. BERT is now Google’s answer to this development.

How does BERT influence SEO optimization?

The short answer: Those who write for users have nothing to fear – BERT only wants to better understand search queries. Nevertheless, it matches these queries with websites and ranks them accordingly. The goal: to find the right answer, and as quickly as possible. Users do not want to scroll unnecessarily, search, or reread; they want a quick solution to their problem. Therefore, one should write for these users – clearly, concisely, and understandably. Those who stuff their texts with fancy keywords and write only for the algorithm will lose out.

Nevertheless, BERT could lead to less traffic on a website, even if it is high-quality and written for users. The reason: The information provided does not exactly match the search. However, this could also have a positive effect, namely „better“ users. These only come because they are looking for exactly what a page provides – and are more likely to become customers. Nevertheless, this is still a pipedream, because BERT does not yet have a start date for Germany. Those who are looking for further information on SEO optimization will find many detailed videos here – with precise instructions.