7 quick takeaways for Google’s search algorithm BERT

Thursday, January 7, 2021  |

If you’re asking the question “What is Google BERT and how will it impact my site?” our helpful rundown of this novel search engine algorithm that seems ready to shake up the World Wide Web!

1. Google BERT is a machine learning technology used for natural language processing in its eponymous search engine. It uses a transformer-based deep learning model to handle sequential data in the form of the strings of words and sentences found in natural language. It represents one of the biggest changes to the functioning of the search engine since Google’s implementation of the machine-learning Artificial I, RankBrain in 2014.

2. Before you look for his scruffy yellow friend Ernie, BERT actually stands for Bidirectional Encoder Representations from Transformers (BERT). The bidirectional nature of this neural-network technique means that it ‘reads’ groups of words both from left to right and right to left to gain an understanding of what is being said. In this way, BERT is being used to prime Google Search with an enhanced capacity for understanding what users are looking for.

3. Google BERT was first introduced in October 2019 for English language queries and since then it has expanded to over 70 languages globally. Initially, the multilingual focus of BERT was focused on featured snippets but now the language processing capabilities will be used to better match results of queries to user intent across the globe.

4. The implementation of this ingenious question answering system has required Google to deploy novel hardware that has the processing power to serve up the newly enhanced search results. Googles Cloud TPU pods facilitate the utilization of their proprietary AI learning infrastructure to deliver search engine results quickly.

5. BERT is expected to impact up to 10% of all searches on Google. This will mean that the impact on Google SERPs is noticeable, especially given the increasing prominence of featured snippets in search results. Googles priority is always the user and they expect that adding BERT to their algorithmic mix will make it much easier to find useful information.

6. BERT works best on user queries that are more conversational or lengthy. In particular, it is able to recognise the importance of “for”, “to” and similar prepositions that can be critical to what is being meant. BERT supports the search engine in understanding the context of the natural language used in queries, building on a prolonged period of extensive testing of this technology. With BERT it is all about the peculiarities and nuances of the way people use words.

7. According to Google, you cannot optimize for BERT. This is all about understanding natural language better. The best focus of your effort will be in providing useful, relevant information to visitors to your site. By paying close attention to Googles Expertise, Authoritativeness, Trustworthiness (EAT) quality standards, you should be producing the kind of content that BERT is likely to digest and rank.

 Are you scratching your head about Search Engine Optimization for your website?

Why not reach out to our friendly and professional team of SEOs who are on hand to answer your questions and share their expertise. Getting in touch is as simple as calling or sending an email. We would be delighted to assist you in bringing more visitors to your site.

0 0 votes
Article Rating
Notify of
Inline Feedbacks
View all comments

Need something specific?

Share with us, We'll build it

  • Your details

  • Your requirements

  • This field is for validation purposes and should be left unchanged.


111 West George Street Glasgow G2 1QX


101 Clerkenwell Road Holborn London EC1R 5BX

    Your details

    Your requirements



    We help e-commerce businesses grow revenue and increase customer reach through the optimal combination of omni-channel presence, user-centric solutions and know-your-customer tools.


    Get in touch today, we’re happy to arrange meetings via:
    Zoom - Skype - WhatsApp - FaceTime or as appropriate


    We’re here for you!