What will affect 1 in 10 search queries? The answer is Bidirectional Encoder Representations from Transformers, more commonly referred to as BERT. Described as the most significant Google update since the release of RankBrain, the Google BERT update represents an improved understanding of natural language in search queries. Prepositions, like ‘for’ and ‘to’ might be little words, but when it comes to search queries, they can make a big difference.
What is Google BERT, and How Will it Affect my Website?
With the BERT update, Google can assess the way words in a search query relate to each other, making it better at identifying nuances and context in a search. This improved understanding of a searcher’s intent is great for users, providing them with more relevant search results. With such a big update the even bigger question for businesses is, how will Google BERT affect my website?
What is Google BERT?
Ten percent of all search queries is a big number, and that’s the estimated number of queries that are expected to be impacted by BERT. Even if your website isn’t affected initially, as your website grows it will eventually affect your website.
The update involves a language processing technology called BERT, which can assess the way words in a search query relate to each other, and should give Google a more nuanced understanding of a searcher’s intent. Simply put, BERT will help computers understand language more like humans do.
How do Google algorithms work?
Google ranking systems sort through billions of webpages to find the most relevant and useful results. These ranking systems are made up of a series of complex algorithms that tell a computer how to complete an assigned task. Google’s search algorithms look at many factors including keywords, relevance, usability, expertise, and location. The algorithms assign a rank to each page, and higher ranked pages will appear further up in Google’s search engine results page (SERP).
How will BERT affect my website?
It is expected that BERT is most likely to affect long-tail searches, which is predicted to be about one in 10 searches in English in the US. Previously in a search query, search algorithms were trained to look at the ordered sequence of words, either left-to-right or combined left-to-right and right-to-left. Meaning the algorithm only looked at the words either immediately before or after a word to search for context. BERT will be able to discern search intent better because it uses word context based on surrounding words.
Google provided a few examples to illustrate BERT’s impact, starting with ‘2019 brazil traveler to the US need a visa’. Previously, before BERT, the top search result would have been how US citizens can travel to Brazil without a visa. However, the original search wasn’t about US citizens traveling to Brazil, it was about Brazilians traveling to the US. This is an example of a small preposition ‘to’ making a big difference in search intent. With BERT, the top result returned for this query is a link to the US Embassy in Brazil for tourists and visitors.
BERT is also impacting English-language featured snippets. Here’s another example from Google. Before BERT, the search query “parking on a hill with no curb” returned a featured snippet that demonstrated how to park on a hill with a curb. The previous algorithm put too much importance on the word ‘curb’ while effectively ignoring the word ‘no.’ With the BERT update, a search user would see a featured snippet demonstrating how to park on a hill without a curb.
Of course, not all search queries are long-tail or conversational, and they don’t all include prepositions. Shorter phrases and specific branded searches are two examples of queries that may not utilize BERT’s natural language processing. At this time, BERT has rolled out for English language queries. However, the algorithm will eventually expand to all languages in which Google offers Search.
What can you do about BERT?
The bad news for businesses is that you can’t optimize for BERT, but that doesn’t mean there aren’t things you can do. Start by focusing on responding to the intent of the searcher. If BERT is focused on providing more relevant search results to users, you want to make sure you are creating that content. Think about satisfying the specific motivation of a searcher and creating the best experience you can for your audience. Track your long-tail keyword ranking performance and note any changes.
Worried about how BERT is affecting your website? Let Vende Social take a look around. When Google makes changes that can affect your search rankings, we pay attention. We have been helping B2B’s develop winning online solutions since 2000. Contact us today to see how we can help you grow your business.
- Google BERT is a language processing technology that will help give Google a more nuanced understanding of a searcher’s intent.
- BERT is expected to primarily affect long-tail searches, which is predicted to be about one in 10 searches in English in the US.
- Businesses can’t optimize for BERT, but there are steps they can take.