How google's BERT algorithm affects your website traffic
BERT-which stands for Bidirectional Encoder Representations from Transformers, a Natural Language Processing technique-is an algorithm that has been added to Google Search to improve the accuracy and relevance of the search results produced. On October 21, 2019, Google officially activated BERT for English-language queries and featured snippets.
BERT collaborates with previously existing Google algorithms to facilitate a clearer understanding of the searches users make by helping the system better recognize the natural language patterns people use. This enables Google Search to identify and deliver more appropriate information-results that better match the user's input. It does not compromise or serve as a substitute for another algorithm but rather builds on the latticework foundation established by previous updates, including Fred, Possum, Mobilege, Pigeon, Hummingbird, Penguin, and Panda.
BERT understands that a single word can be interpreted multiple ways, based on context. For example, in these two phrases,
- "wooden furniture store"
- "store furniture when moving"
The word "store" has two different meanings. Although humans can easily recognize and understand the difference, a search engine would have difficulty doing so.
The focus, therefore, is on longer, more conversational searches in which words such as "to" and "for" are key in overall search meaning.
Google says that for the moment, the BERT update will affect (i.e., help Google Search better understand) approximately 10% of the search queries in the United States.