BERT Algorithm

Google’s BERT algorithm and Search Engine Optimization (SEO) (Part 1):

I know it’s a little technical and most of you reading this article must be thinking, “is it necessary to go into such details if we know how to efficiently use SEO to suit my needs? But I guess with the latest updates rolled out by Google related to the BERT algorithm it’s better to get know a few details about the changes it will make to the current SEO practices.”

So, the first question that comes to the mind is;

What is BERT?

BERT stands for “Bidirectional Encoder Representation from Transformers”

(Quite difficult to understand right?)

Yes, it is a bit complicated to pronounce but its function is simple, Google has rolled out the BERT algorithm to understand the processing of natural language in a better manner.

For Example; you input a train number in Google, and you get its complete status (similar for flights) or if you search for a math equation on Google and a calculator shows up.

The one question after reading the examples that you will have is Google already had these features and the current SEO practices are also in alignment so is there a point of understanding the BERT algorithm at all.

But the intricacies and the minute changes Google has rolled out is what makes it worthwhile to understand.

Let’s take into consideration examples to better understand the usefulness of the BERT algorithm.

Before BERT if someone would have searched “Canada tourist visa for Indian citizen 2019”; then google algorithm would have shown general news articles or white papers (optimized with SEO) as the top results.

Now here is the catch; if you type the same query now, the BERT algorithm considers all the prepositions like ‘for’ or ‘to’ etc which can add to the broader context of the query and make it more relevant. Thus, now after inputting the query in Google, BERT algorithm shows you the most relevant article. In the case of the query “Canada tourist visa for Indian citizen 2019”; it will show you the page of the Canadian embassy providing tourist visa information.

Still confused, let’s make it simpler:

To summarize the above example; the Google BERT algorithm now considers the most relevant meaning and context of the prepositions and the words used in the search query. This means google algorithm now understands that ‘stand’ and ‘standalone’ are separate words with their own contexts in the search query, similarly it has also learned to differentiate between ‘worth’ and ‘worthwhile’ etc. So, the search queries now yield more relevant and accurate results.

(For SEO professionals this means that the words used in their content should be precise and the context of the content should be thorough in order for Google BERT algorithm to read and associate it with the search query).

As I want to maintain a clear understanding of the effects of the BERT algorithm, I will be delving into individual changes to writing content and also to current SEO strategy with bi-weekly blogs.

So, what are your thoughts about this update? Do share your opinions as it will help to increase the knowledge around the BERT algorithm and also in creating effective content and SEO strategies.