SEO Blog 13 min read
What is Google BERT and How Will It Affect Your SEO Strategy
- 31 Oct 2019 6 min read
Google recently announced the rollout of Google BERT, its most significant algorithm update in the past five years. The search engine company announced that the latest Google algorithm update now affects one out of 10 search queries in the US, and has affected featured snippet results globally.
What is Google BERT and how will this affect your SEO strategy? As a pioneer SEO company in the Philippines, Propelrr is here to give you the answers to your most sought-after questions about the latest Google algorithm update.
To help you navigate through this article on Google BERT, here’s a summary of the topics to be covered:
- What is Google BERT?
- How does BERT work?
- Why is BERT significant to you?
- How does BERT affect your SEO Strategy?
- How can you optimize for Google BERT?
- Key takeaways
Ready for a deep dive on BERT? Keep reading to learn more about Google's latest and most significant update since RankBrain.
What is Google BERT?
BERT, also known as Bidirectional Encoder Representations from Transformers, is Google’s a neural network-based technique for natural processing language (NLP) pre-training. Simply put, the BERT update helps computers better understand language the way humans do.
Its application to Google search means the techniques help search engines better understand the reason behind a search; thus showing more relevant results on the SERPs. Here's an example from Google’s BERT article:
The search query is “2019 brazil traveler to USA need a visa.” As humans, we understand it as someone, a Brazil traveler, is asking if he/she needs a visa when going to the USA.
The search results for "before" show how a US citizen can travel to Brazil. That’s not at all what the user intended to search for, according to his query. This leads the user to two situations:
The user clicks on the first result given by Google; but since the article doesn’t have the information he/she needs, the user will quickly exit the page to look for more relevant results. This will then affect a website with a low time-on-page, and a high bounce rate which are not good analytics results.
The user runs another search using a different keyword or just alter the original query. There might not be website analytics involved, but then user experience on Google might not be great since he/she didn’t get what he/she was originally searching for.
With the rollout of Google BERT, the "after" results show just exactly what the user intended to search for. The search engine gave the user relevant and correct information for a Brazilian traveler who needs to know if he/she needs a visa when traveling to the USA. This algorithm update is what changes the SEO game. Google now understands longer tail, natural human language better than before.
The days of unclear search results are almost over, all thanks to Google BERT.
How does BERT work?
Google BERT is built upon recent work in pre-training contextual representations. According to Google AI’s article on NLP, this includes semi-supervised Sequence Learning, Generative Pre-Training, ELMo, and ULMFit. Unlike these previous NLP models, the BERT model is “the first deeply bidirectional, unsupervised language representation, pre-trained using only a plain text corpus,” say Google scientists.
In short, the BERT model is an innovation in language modeling. This is because context is now part of Google’s search algorithm, making it more capable of understanding human language like never before.
Google says BERT helps better understand the nuances and context of words in searches and better match those queries with more relevant results. It is also used for featured snippets, to give users more useful information.
What makes BERT significant to you?
Other than “representing the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search,” it will also be affecting your search results and ranking sooner than you think.
To date, 10% of all queries in the USA have been reported to be affected by this major update. In the digital space, this covers a significant amount of search queries.
Aside from queries, this will also affect the "Featured Snippet" results. Since the algorithm can now read contextually, the search results will also change to a more useful and informational one, at that.
How will Google BERT affect your SEO?
Unlike core algorithm updates such as Google Penguin or Panda, the BERT update will not penalize websites. Penguin went after link manipulation efforts, while the Panda update went after content manipulation efforts. (Trust us, we know what it was like post-Panda update.) This time, Google BERT will aim to improve the search quality of the Google search engine to understand search intent queries better.
The update may also affect your search traffic. BERT is more about understanding more specific longer tail, and natural language queries. Making SEOs generally track them less compared to short-tail queries. Don’t worry about the traffic; it’s actually a good thing.
Sure, you may lose some traffic from this update, but the traffic was ruining your user metrics and increasing your bounce rate. You wouldn’t want that, right?
How can you optimize for Google BERT?
The simple answer is you can't. There’s nothing you need to do or change to “rank” for Google BERT. Just like how you can’t rank for RankBrain, you can also not optimize for BERT. If you want more proof, Danny Sullivan of Google tweets:
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.
Instead of searching for impossible ways on "how to rank for Google BERT," here’s the best SEO advice on how to optimize your site, in accordance with the update:
1. Be clear with your search intent
With Google Bert making "context" matter even more than before, it’s time to revisit your content’s search intent. Are you writing this article to give users valuable information, or are you writing to encourage users to buy a product?
Ahrefs' "Why Search Intent Matters" cites four different types of search intent, namely:
This search intent aims to give users information on what they are searching for.
This is for users who want to know where they want to go or are looking for a website but have no idea what the URL is.
Catering to a buying audience, this search intent is targeted at people who are in the purchasing mode.
This is for users who are weighing their options on what product or services they will purchase. These users have yet to make a decision and are still searching for other options.
Of course, some queries have multiple search intent. You have to gauge which one you want to target; so that everything—from keyword research to content angle and the content itself—is aligned to the overall content goal you want to achieve.
2. Produce better quality content
If there’s one thing you can do to optimize for Google BERT, it’s to write and produce better content for your website. Content is king, after all. The latest algorithm update rewards websites that provide valuable, informative, and well-written content. There’s no room for vague words and sloppy text with BERT.
3. Write for people, not for BERT
Do optimize for humans, not for Google BERT.
Write for your users, your readers, and people who will actually understand your content. Google devised an NLP technique to read and understand content the way humans do. No more “writing for the bots.” It’s all about writing for living, breathing, human beings.
Here are two essential takeaways you should take note of with the update:
- Don’t take the update lightly. It may not be a core algorithm update that affects sites directly, but it will make a big effect on search sooner than you think.
- Create better content. Time and again, content will always remain king. No room for fluff and vague words; get your point across as smoothly and as quickly as possible.
There you have it, your SEO guide to Google’s latest algorithm update, BERT. Is there anything we missed out on? Or do you want a more in-depth discussion? If you do, just send us a message on Facebook, Twitter, or LinkedIn. Let’s talk!