Reading time: 8 minutes
Article Summary: Google search has undergone substantial changes since the rollout of Bert, the latest technology developed by Google’s AI team. For digital marketers, BERT is a real challenge because, even though we reached some conclusions upon certain important factors to pay attention when optimising websites, there is still a lot to discover and test.
From RankBrain to BERT: All You Need to Know About the Update
If you are a digital marketer, I am sure you are aware of the important changes brought to Google Search by the introduction of BERT, Google search’s biggest change since the introduction of RankBrain. BERT update started rolling out on October 24th 2019 and has shown an important impact on rankings for many English language queries.
For SEO specialists, BERT is a real challenge because, even though we reached some conclusions upon certain important factors to pay attention when optimising websites, there is still a lot to discover and test.
We know that, for a busy marketer, staying in touch with all the news in the field may be difficult. This is why, since we have already read a ton of material on Bert and NLP, we would like to share our most interesting findings with you.
What Is Bert Update and what is it used for?
We start by explaining what BERT stands for and how it functions. Bidirectional Encoder Representations from Transformers is Google’s neural network-based technique for natural language processing (NLP) pre-training. The AI-powered Google update was created and published in 2018 by Jacob Devlin and his colleagues from Google and is used to help machines better understand the nuance and context of words in search queries and better match the queries with useful results. In short, if Google used to process words one-by-one in order, thanks to the new technology, Google processes texts as a whole, understanding context and the searcher’s intent behind search queries.
How Does BERT function?
According to Search algorithm patent expert Bill Slawski (@bill_slawski of @GoFishDigital), this is how BERT functions:
“Bert is a natural language processing pre-training approach that can be used on a large body of the text. It handles tasks such as named entity recognition, part of speech tagging, and question-answering among other natural language processes. Bert helps Google understand natural language text from the Web.”
Bert vs Rank Brain.
When RankBrain was introduced by Google in 2015, it had the same purpose as Bert: to improve users’ search intent and deliver them more useful results. What RankBrain did is analyze both web content and users’ queries in order to understand the relationship between words and the context of the query. Actually, the Bert update is complementary to the RankBrain algorithm and it doesn’t replace it completely. Google may decide how to interpret a specific query and it may apply one technique or the other, or sometimes even a combination of the two.
Here are some examples of queries’ results before and after the algorithm change offered by Neil Patel:
We can clearly notice that the second result better matches the searcher’s intent. The result offered by Bert is more relevant in every example.
Why Is Bert Useful?
Applying BERT models to both rankings and featured snippets in search, Google is able to offer its users more useful information. When it comes to ranking results, BERT will help machines better understand one in 10 English language searches in the U.S. The algorithm will be extended to more languages and locales over time.
What is NLP (natural language processing)?
Natural language processing (NLP) is an AI-powered technology which helps computers understand human natural communication. The objective of NLP is to read, understand, and make sense of the human languages in a valuable way. Most NLP techniques rely on machine learning to derive meaning from human languages. NLP is not a new feature offered by search engines, but Bert represents a breakthrough in NLP due to the bidirectional training (instead of analyzing and training ordered sequence of words, Bert analyses and trains language models based on an entire set of words in a sentence or query).
To sum up, NLP is the process of analyzing a text, understanding the meaning of the words and establishing relationships between them, in order to interpret them more naturally.
Major components of NLP
The key to well-optimized content for NLP is a simple, clear sentence structure. To do a good SEO & content marketing job, you need to have a basic understanding of NLP’s core components:
Sentiment can be defined as the score of the sentiment (view or attitude) about entities in an article.
Named Entity Extraction: entities are generally people, places, and things (nouns). Entities can also include product names – generally, the words that trigger a Knowledge Graph.
Tokenization is the process of breaking a sentence in separate terms, while Parts-of-Speech Tagging classifies words by parts of speech.
Through lemmatization, Google can determine if words have different forms, and word dependency creates relationships between the words based on rules of grammar.
With the help of Subject Categorization, NLP classifies text into subject categories, like Arts & Entertainment, Adult, beauty, Law& Government etc.
How We Optimize Content After Bert
In October 2019, the famous Google search engine liaison Danny Sullivan claimed in a Tweet post that there is nothing to be done to better optimize content for Bert.
There’s nothing to optimize for with BERT, nor anything for anyone to be rethinking. The fundamentals of us seeking to reward great content remain unchanged.
According to Neil Patel, Bert is “mainly impacting top-of-the-funnel keywords, which are informational keywords”. Many other SEOs noticed that long-tail expressions and conversational queries are mostly impacted by the update.
Therefore, if you want to maintain good rankings and also beat the competition, you should try to get very specific with your content. And, while many SEOs may advise clients to write super long content, the truth is algorithms mainly focus on content quality, and not quantity. Also, keyword density is not important anymore, because Google takes into consideration the context the keywords are used in, rather than the number of repetitions.
Using different tools like Buzzfeed, Buzzsumo, Quora or Reddit to find new topics for website/blog content is a great idea, but content should be super specific to be able to rank. For example, if you write an article on “The best way to lose weight without diet or exercise”, you should really focus on many alternative weight loss methods, without writing about any type of diet or exercise type.
Actually, the best way to optimize for Bert is to write good quality content for your targeted audience. By increasing your website’s relevance you also increase its chances of getting better rankings. In its press release from October 24th 2019, Google claimed this big change is “representing the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search.”
To rank better than the competition for a specific keyword, you don’t have to create a 5000-word blog post dealing with 100 different topics. It’s more useful to create a unique, detailed piece of content answering a searcher’s question and providing more value compared to the competition.
Another good search engine optimisation idea is trying to deal with secondary topics in your blog articles. If the initial target keyword is a broad, head term, you can treat many secondary topics. This way, you’ll naturally use additional expressions that improve salience. But if you optimize for a specific longtail keyphrase, you should stick to the main topic only.
The structure and formatting of the content are very important for ranking. To help the algorithm understand your content and consider it clear and useful to the users, make sure you use the following: headings, subtopics, HTML tags, ordered lists, inverted pyramid structure.
Targeting featured snippets, paying attention to content readability, focusing on thorough topic research, using checklists and Q&As may prove successful content marketing techniques in the Bert era.
Even if you lost some traffic due to the latest update, this is not necessarily a bad thing. Maybe the content you produced attracted fewer visits, but if the bounce rate decreased and the time on page has increased, this probably means that the users who checked your piece of content found exactly what they were looking for.
To sum up, in order to consider the new algorithm a threat, we should all focus on the opportunities that Bert offers us to create better, more complete, super-specific content to answer our targeted public’s questions.
In the next video, Search Engine Journal specialists thoroughly explain how Bert functions and what are the implications for marketers :
Also, find below an infographic explaining and summarising all BERT’s main characteristics.