• Prosenjit Ghosh

What is NLP or BERT in SEO and how can you use it to sky-rocket your organic traffic ?

Updated: Sep 1, 2020

Audit your website for FREE today :Fill this Form to rank your site high.

Before starting, let's tell the full forms of both terms. NLP stands for Natural Language Processing. While BERT means Bidirectional Encoder Representations from Transformers. It is not an easy task to teach machines to understand how we humans communicate with each other. That's why scientists and engineers all around the world work on new technologies that could help them understand not just the words but also the real intentions behind them.

But why do companies like Google even bother to develop those technologies? Google's mission is to help people find the most relevant information and that's the core of their existence. To show the right content at the right time, they have to understand the real intention behind the query. This is why Google has been working for so many years on techniques for NLP. Finally, in 2018 Nov, they developed BERT. If you had been involved in the SEO industry in Oct 2018, you probably heard about the algorithm CORE UPDATE. It was one of the biggest updates in the last 5 years.

So what stands behind this mysterious algorithm. As opposed to the directional model like we read texts sequentially from left to right or right to left. The transformal encoder reads the entire sequence once at a time. These unique characteristics allow them to land the context of words based on their surroundings. And that makes the BERT so powerful. Right now I will focus so specifically on SEO.

You will be able to use all these things for FREE with Google tools. Google's Natural Language Processing API is available for everyone and can be used for entity recognition, content classification, and sentiment analysis. What are those elements?

Entity Analysis: It provides information related to the entities in the text. Generally, these are the words or phrases which can be identified, classified, and categorized by Google. Examples of objects are famous people, events, numbers, organizations, etc. Because Google spends a significant budget on classifying them. We can assume that it's an important hint for the algorithm. For example, the word evening is much more important than morning when we talk about dinner.

Sentiment analysis attempts to determine the overall characters of the text: positive, neutral, or negative. The positive sentiments mean the topic has been described favorably. It has positive words like great, guru, hero, outstanding, etc.

Negative sentiments suggest the usage of detrimental statements in the contents. As you probably have already guessed. Those pages have words like hate, weak, stubborn, danger, etc.

As you can see, Google itself separates important words and phrases that we can use for our topic. So using them in our content will increase their relevancy in Google's eyes. There are two ways to get the important list for the words and phrases that you should use in your text.

First, find your own competitors who rank high on Google. if you are writing an article make sure your competitors also have a similar article. Go to the Google NLP API demo tool. Then copy and paste content from your competitor's page to google's NLP API demo. Copy entities with the highest sales curve and paste them into a spreadsheet. Do the same thing with other competitors as well to get better results. Now you have a list of relevant keywords that you should use on your content.

For the second way, I will use the surfer and the NLP feature. It uses Google's API to optimate and speeds up the processing that I described before. Go to the content editor tool, provide your target query, pick the location where you want to rank, and turn on the NLP. Customize the link of the competitors that the server will include in the analysis. Pick only the ones with the highest content score. If they are scored well then they must have done their own page homework and they will be your best benchmark for the analysis. A page with a score below 60, is not a good point to give preference.

Keywords and phrases with a blue badge are the entities scribes directly from Google's NLP API. Save changes, paste your existing content, try writing from the scratch or share a link to the interactive editor to your copywriter.

The second thing you need to do is sentiment analysis. If your content targets the wrong sentiment, It can be a hand-brake for a position. If they suggest only the positive sentiments you can assume that's the way to go. If their sentiments are rather negative, you may focus on improving the downsides in your content. To make you use of the sentiment score, you can compare your site to the other sites in SERB and look for the co-relations. Of-course co-relation is not cognition. But it still can be an important hint. To find out what the sentiments of the SERB you can :

Use the spreadsheets and Google NLP API demo. Just paste content from the sites and then copy the sentiments to get results. Paste it to the spreadsheets and create a chart. Do it for all the pages from the top ten. Or you can speed up the process and use SERB to SERB analyzer. Pick your keyword and location, turn on NLP sentiment analysis. Wait for the query to load and open it. On the left, there are more than 500 factors that surfer checked for your keyword. On the top the NLP sentiments that you can turn on to get the sentiment to analyse within a second.

If the score value is above 0.25, it indicates positive sentiment. As you can see the sentiments on pages that rank for a particular term is definitely positive. We could assume that writing a negatively saturated content won't take us closer to ranking high for the term. But is that worthy of all your time and effort.

Well, some time ago I had a conversation with Mathew James who told me that NLP impacted his processes. So I asked him if he would like to share some results of this. And he agreed. Mathew told: " we have done surfer based NLP reports from the process of content creation and optimization processes back in January 2020 which delivers great results. Whether it's a new site or an existing site that we have optimized with a surfer to quicken the SEO process. If you are not building or optimizing your content which surfer is recommending, you're leaving a lot of search traffic on the table.

So, I hope that henceforth you will have an idea of how you could incorporate NLP to your SEO activities to take advantage of the new, exciting technology.

Audit your website for FREE today :Fill this Form to rank your site high.

Contact :




India- 560066


Mob: 9480119503

E-mail :  support.virtuallly.com
  • Instagram
  • Facebook B&W
  • Twitter B&W
  • LinkedIn B&W
  • YouTube

© 2020 by Virtuallly.com | All Rights Reserved. We work virtually.