BERT, The latest and most significant Google update since Rank Brain is already in effect since one week, and as per Google fellow and Vice President Search, Mr Pandu Nayak, ”It will effect 1 in every 10 google searches”, especially the long tail searches where prepositions such as “for” and “to” have a significant meaning in the sentence. In his own quintessential Googler style, Mr. Nayak writes in the blog,
“At its core, Search is about understanding language. While we’ve continued to improve our language understanding capabilities over the years, we sometimes still don’t quite get it right, particularly with complex or conversational queries. (click here to read full article)
Effects of BERT Update
Experts are terming BERT ( Bidirectional Encoder Representations from Transformers) as one of the biggest Google update in last five years and more importantly a huge step towards how searches will be understood by engines in future. Although it is going to be extremely difficult for SEO managers to predict the impact of applying deep machine learning concept that is pre trained for search queries.
Even if you are Chess Grandmaster!
So keep in mind that you cannot optimize your website or blog for deep learning NLP. All you can do is to and carry on writing quality content for humans and not machines.
But there’s good news too.
It is highly unlikely that your Moz stats or DA rankings of your websites will be impacted because these metrics track only the occurrence of shorter (1-2 word) queries, whereas BERT is deployed to decode longer or conversational queries.
Is this something new?
Yes and No. Though Its effect and integration to search queries will be something new to be felt by SEO managers in India and abroad but work on BERT NLP algorithm has been going on for some time and in fact a NOV 2018 research paper on Google AI blog hinted that it will soon come in practice, not as a substitute for RankBrain but working in tandem with it. BERT technique is being used in other areas such as investing too although Google open sourced it last year.
Enter THE B.E.R.T.
Exaggeration aside, BERT update is already impacting millions of Google searches made using US English since 25th October and also for Featured Snippets in other local languages such Hindi and Korean among others. Soon it will be used to interpret user intent in other locales and languages too.
Meaning! Google aims to apply BERT ( Bidirectional Encoder Representations from Transformers) technique to NLP (Natural language processing) which enables to pre train the system on natural language skills while understanding conversational and long sentences, thus the capabilities of Google search are amplified to better understand the user intent as well as the context in which the supporting words are used in a search query.
Let’s take an example of common search query like Bank. For some Google users, it may mean a bank savings account and for others it may represent a bank of a river. With BERT’s bidirectional reading powers, it will fetches results on a query such as Bank by connecting it to the other words used in the query i.e. left and right values along with the understanding of applied context.
After BERT Effects
In the realm of search queries and SEO practices, an update which affects more than 10% of English search queries is not subtle. It is in fact so huge that some of us have already felt the heat and most of us from the SEO community will closely watch and adapt to this neo technique that uses neural network based approach for Natural Language Processing (NLP).
RIP SEO. Long live Google
Essentially using BERT algorithm to NLP is a kind of cheating because it means that the system already knows the answer before you even ask it. A big SEO Lobby which is getting bigger argues that the real intentions of Google behind applying mathematical solutions to search queries may kill the traditional SEO as we know it. It may lead to user getting all required information on GOOGLE itself and SERP visibility of your website will cease to bring traffic. Meanwhile Google is already claiming that 10-15% search queries in English are better understood since the launch of BERT. All you can do is to continue writing clear and informative content that does full justice to user’s query with unique information and steer clear from keyword clusters that rely on the “Keywordease” concept.
The long term effect however remains to be seen and closely watched.
Should you panic?
Yes, if the content strategy is crawler based and not focused on delivering informative content to your users. For the rest of us, the focus on rich, user friendly content and its rewards remain unchanged.
How can you REALLY optimize your SEO for B.E.R.T.
Write for humans and we cannot emphasis more than that, In the last 11 years, I have worked with numerous SEO teams and strategists who were at different points of the learning curve but only a tad few of them evolved with GOOGLE in terms of putting users first over search bots. The result is a web page/blog that focuses on keyword-density and H tags but misses out on the holistic purpose of providing unique content to the readers. We at Digiversal, couldn’t have been more delighted with BERT update as it endorses and rewards our strategy of investing heavily in top quality content and copy writing for our Indian and International clients in USA, UK etc.