Everything You Want to Know About BERT and Some Things You Probably Don’t.
Just last week, Google announced the rollout of an update in how it handles search queries. This update is known as BERT, and it’s just another example of how Google is constantly upgrading their algorithm to better understand users’ search queries.
Who What is BERT?
BERT stands for “Bidirectional Encoder Representations from Transformers,” which Google describes as “a neural network-based technique for natural language processing (NLP) pre-training.”
Don’t let that jargon-loaded tech-speak scare you. This is all a very fancy way of saying BERT is a way to help Google’s search algorithm better understand language and, more importantly, the relationships between words.
Still Confused About What BERT Is?
Think of it this way–the average Google user (and search engine user, for that matter) tends to search in very conversational terms. It used to be that people would search with short, simple queries, like “pizza philadelphia.”
As search engines evolved and became more prevalent, users began searching with much more conversational queries, like “Where is the best pizza in Philadelphia.”
With voice search becoming much more popular on both mobile devices and home devices (like Google Home and Amazon Echo), search queries have become even more conversational with searches that may look more like “What pizza places in Philly are still open this late?”
For the most part, Google’s search algorithm has been able to keep up with the way conversational search has been evolving. Hummingbird, in particular, was one of the search engine’s first big steps in embracing conversational search.
What makes BERT different, though, is how it allows Google to understand the differences and relationships between different words.
Google has provided some great example of what BERT looks like in action, such as this before and after image for the search “2019 brazil traveler to usa need a visa.”
In the past, Google’s algorithm would have had a difficult time understanding the importance of the word “to”, if it paid any attention to it at all. But in a post-BERT world, “Search is able to grasp this nuance and know that the very common word “to” actually matters a lot here, and we can provide a much more relevant result for this query,” according to Google.
How Does BERT Affect Your Business?
Good question. But because BERT just began rolling out a few days ago (and may not be finished by the time this is posted), it’s way too early to tell for sure.
That said, Google has stated that this update is likely to affect 1 in 10 searches in the U.S. in English. That’s a pretty big chunk of the search results. By contrast, most algorithm updates tend to only affect a very small percentage–often in the 1 – 4% range. That makes this 10% much more substantial.
How Can You Optimize for BERT?
Um, you can’t. Like RankBrain before it, Google has made it clear that there isn’t anything in particular you can do to optimize for BERT.
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.
— Danny Sullivan (@dannysullivan) October 28, 2019
That doesn’t mean you’re completely helpless, though. Even though you can’t do anything to directly optimize for it, you can still continue with what you should have been doing in the first place: Creating comprehensive, easy-to-understand content that will anticipate the questions your customers may have.
Free Tip of the Day
And that brings us to some free advice: Create great content on your site! More importantly, make sure your content covers all the potential issues someone interested in your services/products may encounter.
Selling couches? Don’t just explain the specs of each couch. Instead, go deeper and explain what design setting each couch may best fit in, what type of home, what type of accent pillows will match best with it, etc. Try to brainstorm every possible topic that may be on your customers’ minds and attempt to answer them. Because no matter how trivial those questions may seem, customers are still entering them into search engines, and those search engines are desperate to give them the right answer.