How Google BERT can affect your website

Google rolled out the most impactful update in the last five years – BERT (Bidirectional Encoder Representations from

Transformers). Google BERT is a revised search algorithm update that applies a neural networking approach to ascertain the purpose of the search queries better. This new update is a deep-rooted learning algorithm similar to natural language processing.

It will enable a machine to understand the context behind the words in a sentence. For e.g. look at the picture below where a search query is put forward by someone who wants to travel from Brazil to the USA and needs a visa. Earlier before the update, google would have sent in an answer similar to the first picture. But with this new update, Google can now understand the context behind words better and send a correct answer similar to the one in the second picture.

According to Google, this new update will affect 10% of search queries in the United States, as for the featured snippets the update is up and running on a global spectrum. This new update will enable google in understanding extended and wordy searches/ queries. This comes in a play where simple prepositions like ‘from’ or ‘to’ change the meaning of a sentence completely.

Google BERT will significantly affect websites that don’t have well-curated content or abrupt content leading to ambiguous conclusions. The websites will have to up their skillset in writing and be more definitive in context.

On the flip side, websites that have reliable and sincere content will benefit vastly from this update. As it’s more about breaking through the noise of irrelevant answers and getting to the very niche of what the searcher wants to be answered. It’s not about writing a thesis on 20 topics, it’s about the brevity in which you provide answers to queries.

To keep your position intact and not be affected by this new algorithm update you first need to understand the 3 main types of queries that people generally search. They are: –

  1. Informational
  2. Navigational
  3. Transactional

Informational Query: – It can be searched by someone who wants to cook, learn facts about things, etc. Example of an informational query: – “how to cook Thai curry?”

After performing the search, they may encounter several different recipes for Thai curry. From there they’ll land on a navigational query like ‘Gordon Ramsay’s Thai curry recipe’

Subsequently, on finding the solution they may perform a transactional search query such as ‘Gordon Ramsay’s guide to Asian Cooking’

At this point, BERT is mainly impacting the top funnel keywords which are the informational keywords. A simple solution to maintain your rankings is to be specific and definitive with the content.

Additionally, the focus needs to shift from keyword density. The reason being this new update is facilitating Google to get a better understanding of words and context.

In conclusion, it’s clear that this update will affect websites with unclear content but it’s also a ray of hope for development. E.g. If someone looking for Thai curry recipe lands on a website that shows a chicken curry recipe will anyway drop back to google which will, in turn, hamper the website’s user metric and increase the bounce rate.

So, the only way to overcome this new update is by adjusting the content as per the query. However, changes in SEO to survive the algorithm update should be avoided as the new update will concentrate on understanding human language and will recognize high-quality content and articles through its machine learning.


Published by: cwt