50% of all searches will be voice searches by 2020, and Google’s taken notice with update BERT. Devices are being treated like companions, giving rise to conversational search queries and, with that, content that is written to match the new conversational nature of search engines. Content formats like FAQs will benefit from this update as the questions asked throughout a piece will match the same questions users will ask Alexa. It only seemed obvious us to then use this very format to describe Google’s newest update. Here are the top 10 FAQs surrounding BERT.
What is BERT?
According to Google, BERT “represent[s] the biggest leap forward in the past five years, and one of the biggest leaps forward in the history of Search.”
What does BERT stand for?
BERT stands for: Bidirectional Encoder Representations from Transformers.
Wait, what? Essentially, the update is a result of a breakthrough in Google’s research on transformers. Transformers work to process words in relation to other words in a sentence, rather than processing each word one-by-one. This helps Google to understand the context of the search query, rather than relying on the dying language of keyword-ese.
What is keyword-ese?
Why is context important in search queries?
Voice search is becoming the new norm, making search queries more conversational. Queries resemble sentences instead of keywords. As a result, the individual words themselves are less important; rather, it’s the meaning of the sentence overall that matters.
Who is most affected by the BERT update?
According to Google, it’s expected that 1 in 10 searches will be affected by BERT. Worth noting is that the update analyzes search queries – not content. How Google will then interpret those searches will be what changes the results. Consequently, the traffic you may lose to BERT is likely low-quality and not meant for you anyway. Any traffic you gain will be more relevant and, accordingly, more likely to convert.
How do I optimize for BERT?
Simply put, create high-quality and highly relevant content. Don’t hide behind keywords. Instead, focus on creating content that users want to read. Do this by researching the topics often searched by your target audience and then creating a content calendar around these same topics. This is exactly how inbound content is meant to work – offer something useful to the users you care about, and Google will be more likely to show that content to them.
Good content formats in response to BERT include: how-tos, FAQs, and Q&As. When drafting an FAQ or Q&A, for example, incorporate questions users are likely to search for. Google often uses blogs with these formats for its featured snippets.
Can I see a Google BERT example?
Google included examples demonstrating the effects of BERT in its announcement. One of the examples compared search results before and after the update for the query: “2019 brazil traveler to usa need a visa.”
The intent of the search query is to find results related to visa information for Brazilian travelers coming to the U.S. Initially, the search engine misinterprets the meaning of the word “to,” causing the search engine to display results on U.S. citizen travel to Brazil. After the update, the engine understands the true intent of the search and appropriately displays information on U.S. travel visas.
How does BERT compare to other Google updates?
Two updates keep coming up as the search community discusses BERT – Hummingbird and RankBrain.
What’s the difference between BERT and Hummingbird?
Hummingbird was Google’s first response to the trending takeover of voice search. BERT is the latest. The updates differ, however, in how they process and understand language. Prior to BERT, search queries were understood by reading from left-to-right. Now, it compares words in relation to other words within a sentence, making the order of these words less important.
How does BERT compare to RankBrain?
Many are saying that BERT is the biggest update since RankBrain launched 5 years ago. RankBrain was Google’s first AI used for search queries. Like Hummingbird, it was one of the earliest updates shifting the focus from keywords to natural language. BERT isn’t meant to replace the work of RankBrain or Hummingbird; rather, it’s meant to build on it.
Making the Most of BERT
Any good marketer has nothing to fear as BERT takes effect. If you’re creating top-notch and highly relevant content on a frequent basis, you’re doing exactly what you need to. Two key takeaways: (1) BERT analyzes search queries instead of web content, and (2) the context of searches is becoming far more important than keywords.
Have you noticed any changes on your site due to BERT? Let us know in the comments.