The BERT algorithm is Google’s latest pursuit to improve their understanding of search intent. For eCommerce businesses, the unique details found on their product pages could now drive more organic search traffic with higher purchase intent.
Google’s VP of Search, Pandu Nayak, called this update “the single biggest change we’ve had in the last five years and perhaps one of the biggest since the beginning of the company”, with it set to impact one in ten search queries. The SEO landscape is ever changing but the quest remains the same: to bring high quality, relevant content to users. Over the past five years, we’ve become accustomed to working with Google’s RankBrain but now, BERT helps to understand searches better than before.
What is BERT?
BERT stands for Bidirectional Encoder Representations from Transformers and acts as Google’s Artificial Intelligence (AI) centre for Natural Language Processing (NLP). Like RankBrain, BERT uses AI machine learning to help Google interpret the context of long-tail search queries and provide relevant results. The update is capable of leveraging context, helping computers understand language at a human level. Even if a keyword isn’t present in the search query, BERT can still deliver higher quality, relevant results.
RankBrain and BERT: What’s the difference?
RankBrain, which launched in 2015, is Google’s machine-learning AI system used to help process search results. Some of BERT’s capabilities may sound similar to Google’s first AI method, however, their processes differ. RankBrain adjusts results by looking at the current query and finding similar past queries to display relevant results. It also helps Google interpret search queries so that it can display relevant results that may not contain specific words. The example below shows how Google was able to figure out the user wanted to know specific information on The Shard, despite the name of the landmark not being used in the search query.
Google has reported that 15 percent of all search queries are new, meaning they are being searched for the first time. As a result, a more comprehensive, effective method was essential to help understand all search queries. This is where BERT comes in.
Although it is an NLP algorithm, BERT operates differently. Typically, NLP algorithms consider only one word at a time, whereas BERT can read a full sentence while also considering the individual words that make up the sentence, understanding how those words relate to one another. This is a crucial advancement in NLP as the human language is naturally complex.
BERT does not replace RankBrain. Instead, this algorithm update is a step-up in the same area, designed to be used in tandem with RankBrain to decode search intent and web page content to produce even more relevant search results.
BERT and eCommerce
The BERT update is good news for the eCommerce industry as it will help rank products and category pages higher that meet search intent. Product pages and filtered product grids tend to rank better for long-tail queries; these pages now have a higher potential to rank in SERPs, helping drive quality, organic traffic.
How this algorithm will affect your online store depends on the quality of your website content. If you’re already producing high-quality, authoritative and relevant content, you shouldn’t have much to worry about. However, your product page details are an important consideration with BERT with the update providing an opportunity to optimise product descriptions with unique, detailed and relevant information. Product information such as different types, colours, sizes, features and descriptions have a higher chance of being featured within SERPs thanks to BERT.
When shopping online, users make their purchase decisions based solely on the information provided to them which means high-quality content is essential. When writing product content, you should consider:
- Writing for the visitor, not just to sell
- Using your own product images rather than that of the manufacturers
- Optimising the descriptions of your products with relevant keywords
- Providing visual cues for shoppers and text labels for the search engines when adding attributes such as colours
BERT will impact both organic results and featured snippets, altering rankings to produce more relevant and specific results to match user intent. This algorithm update is likely to increase traffic to some web pages and decrease traffic to others. Pages that see an increase will be more relevant to the search query and should perform better.
Google’s search algorithm is still working to understand the nuances of human language, but the BERT algorithm is yet another step in the right direction. With its ability to understand context of an entire query, BERT provides eCommerce businesses with an opportunity to drive more relevant search traffic to their product and category pages. Are you interested in learning more about our eCommerce SEO services? Contact us today.