putting our heads together

e: [email protected]

t: 01142 509 578

what is bert and how it will affect seo? banner

what is bert and how it will affect seo?

You may have heard about the latest Google update, BERT. If not, this article will take you through Google’s new ranking factors.

 large supporting image

Google are renowned for their frequent algorithm updates, as well as larger core updates. This creates an ongoing challenge for marketers and SEOs. They need to constantly monitor updates and optimise content in line with the latest algorithm changes. This makes BERT a welcome change for SEOs. Its main focus is on relevant and informative content – as long as content creators adhere to this, there should not be much to change.

What is BERT? 

BERT stands for 'Bidirectional Encoder Representations from Transformers'. Google have been quietly rolling out these changes to increase the search engine’s understanding of natural language processing.

Many people may have missed this update because it does not directly affect the tools SEO managers use, such as Moz or RankRanger. These tools tend to track short phrases and keywords, whereas BERT aims to understand longer-tail keywords, which are more natural to human language.

How does BERT affect SEO?

While BERT will not affect the majority of search queries, SEOs should not ignore it altogether. Google says BERT will only be used in 1 in 10 searches, but this is largely because it is so advanced. At present, the update pushes the limits of what Google's hardware can handle. 

However, SEOs should be wary that Google may expand its hardware capabilities in the future. This means that BERT could potentially affect more searches. SEO agencies and search marketers should start implementing changes now to prepare for these updates.

The BERT update will also affect content length. Previously, SEOs would recommend that longer content performed better in search rankings. This is no longer a priority for Google. Longer content is still valuable, as it is designed to be broad and capture as many potential search queries as possible, but now, Google is looking for specific answers. Marketers should create content with specific user queries in mind.

What to do?

So, how can you adapt your content to optimise for BERT? As mentioned before, if you’ve already been following best practices, there shouldn’t be much for your SEO agency or marketing teams to change.

Instead, it’s more a case of focusing efforts on being specific and relevant. All of Google’s updates have the same common theme: making content more useful to the reader. Therefore, the best advice as always is to focus on optimising content for the end user, rather than thinking about how to tick boxes for Google.

Specifically, the following are great ways to make the most of the new updates:

Focus more on informational queries, less on the navigational and transactional queries

As per the marketing funnel, there are three types of query people use – usually one after the other. Informational queries come first. These are usually broader questions such as “how to get more sleep”. These broad queries will bring up a list of potential options.

Typically, a user then will home in on a particular answer and make another search for this. Let’s say “herbal sleep remedies” came up as a result for “how to get more sleep”. The user would then search for this, which is known as a navigational query.

Finally, the transactional query is the step towards a conversion. The user has decided exactly what he/she needs and is searching to transact, e.g. “buy lavender online”.

BERT encourages SEO agencies and search marketers to focus on the informational query more than ever before. This is because Google want their users to be completely informed before they make a decision. If they can streamline this process and get an answer in a single query, even better.

BERT is designed to allow Google to better understand a user’s intentions from that first informational search, presenting them with the most relevant article to solve their problem.

Create more content related to specific questions

BERT encourages us to think about the specific questions our target audience might be asking, and focus on how to answer that specific question.

Therefore, the best piece of advice would be to make sure you think this way when you create your content. Take the time to focus on what the reader actually wants to know, rather than simply casting the net wide in hopes of more traffic. More specific results may result in lower traffic, but it will be higher-quality traffic. It will reduce your bounce rates and, crucially, increase your chances of conversions.

To summarise, SEOs and search marketers don’t need to go making drastic changes, providing they are following best practices. However, they need to be aware that BERT will affect more and more search results in the future.

As with all of Google's updates, no one truly knows everything there is to know, except for Google. Likewise, we never know what’s coming next. It’s always best to focus on the end user experience, as Google's priority with any update is to make the user’s experience better and easier.

Therefore, making your content engaging and relevant is the perfect way to make sure you rank as highly as possible.

How it affects searches and results 

So how does BERT affect searches? The BERT model is said to affect 10% of search queries, giving a more accurate result in the search engine results pages. BERT aims to pick up on keywords such as “to” and “from” to fine-tune its results. For example, users searching for a flight will see the correct order of their origin and destination.

As well as this, searches will perform better with negative words such as “no”. For example, a user may search:

“Picking up a Royal Mail Parcel with no ID"

Rather than prioritising tips on how to collect a parcel from Royal Mail, Google will now show more relevant answers such as, “I've forgotten my ID, can I still collect my parcel?”. Before BERT, these results may have ranked much lower. Google now understands that these are far more relevant to the user’s actual query.

think hydra would be the solution for your next project?

think hydra would be the solution for your next project?

then what are you waiting for, let's start a conversation

Contact Us