Google Update Timeline
So, what’s the deal with BERT?
Google’s latest algorithm update has many marketers and SEO pros wondering how the search giant’s recent changes will impact their brand and outreach efforts in the months ahead.
We’re here to help. Below, we take a closer look at what BERT is, how it affects Google search and what the BERT update ultimately means for your digital marketing team.
What is the BERT algorithm?
Short for Bidirectional Encoder Representations from Transformers (say that three times fast!), the BERT algorithm update was rolled out in October 2019 to help Google better grasp the context of words people enter into the search bar.
In other words, Google is using BERT to better understand the true meaning of words within each unique search query. In turn, this will allow Google to better grasp nuanced phrases and deliver more relevant results to the user.
For instance, the word “imaging” appears in both “photo imaging” and “brand imaging,” yet in each case “imaging” means something quite different. With BERT, Google equips its algorithm with the ability to not only distinguish between the two, but to populate SERPs with results better aligned with what the user is after.
Touted by Google as “one of the biggest leaps forward” in Search history, BERT is designed to dissect and understand the intent of user searches—something that may have a significant impact on how you build brand visibility, relevance and authority down the road.
How does BERT work?
Even though many in digital marketing and SEO (us included) refer to BERT as an update, it’s more accurately described as a technique, one based on a branch of artificial intelligence known as natural language processing (NLP). NLP is centered on the science of linguistics and provides computers the training to better understand how and why humans communicate the way they do.
NLP itself isn’t new in the world of search technology. But BERT does represent the next step forward in NLP—a breakthrough that continually trains Google to consider the entire group of words in a query rather than just the ordered sequence (i.e. left-to-right).
Through “bidirectional training,” BERT enables Google to learn word context within the entire phrase rather than just looking at the words immediately before and after it.
Google’s Pandu Nayak examines how BERT’s “deeply bidirectional” capability goes further to determine context and unearth search intent:
“Here’s a search for “2019 brazil traveler to usa need a visa.” The word “to” and its relationship to the other words in the query are particularly important to understanding the meaning. It’s about a Brazilian traveling to the U.S., and not the other way around. Previously, our algorithms wouldn’t understand the importance of this connection, and we returned results about U.S. citizens traveling to Brazil. With BERT, 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.”
As Nayak notes, even a seemingly straightforward word like ‘to’ can have widely different meanings within different contexts. And by considering how each word fits within the unique context of each query, BERT allows Google to answer questions and deliver results with more precision than ever.
How does BERT affect Google search results?
It’s estimated that BERT improves Google’s understanding of about one of every 10 English-language searches made in the United States. Given that over two trillion searches are performed on Google each year, this amounts to more than 200 billion queries and featured snippets that will likely be affected by BERT in some way in 2020.
While BERT is currently applied to English-based searches in the U.S., Google has plans to apply BERT’s advanced NLP to more languages and locations across the globe in the future.
“A powerful characteristic of these systems is that they can take learnings from one language and apply them to others,” says Nayak. “So we can take models that learn from improvements in English (a language where the vast majority of web content exists) and apply them to other languages. This helps us better return relevant results in the many languages that Search is offered in.”
According to Nayak, a BERT model is now also being used to enhance featured snippets in the 24 countries where it is available.
Because BERT is designed to improve Google’s understanding of more nuanced searches—such as those with words like ‘to,’ ‘for’ and others that can have a dramatic impact on context—its influence will be much greater on long-tailed search phrases than on shorter queries. For example, a search like “where do I go for great dinner recipes” is much likelier to be affected by BERT than, say, “great dinner recipes.”
Of course, all of this is Google’s estimation of what BERT will reach and how it will affect search queries. That doesn’t mean it’s wrong or way off base; just that there are always unintended consequences marketers and SEOs should look out and prepare for when adapting your digital outreach strategy.
How does BERT affect SEO?
Even though BERT is said to affect only 10% of all queries, it will still have a considerable impact on long-tailed search phrases (as we explored above). Since longer search queries are more natural to language and so important to SEO, your marketing team should take care to account for BERT when designing your brand imaging and content development efforts.
In other words, BERT is sharpening Google’s ability to provide specific answers to long-tail queries, meaning you’ll likely need to refine your marketing approach to keep up. Thinking of your content marketing strategy as a way to address your customer’s specific concerns is a good place to start.
Also, just because BERT focuses on long search phrases now doesn’t mean Google won’t apply the technique to different or shorter searches down the road. SEOs and marketing teams need to be aware of this possibility and develop a mindset of readiness so as not to be caught off guard.
How to optimize SEO for the Google BERT update?
Optimizing for the BERT update doesn’t mean making a massive realignment of your current outreach strategy, but rather refining your approach to deliver content that’s more relevant and specific to what your customers are looking for.
While Google’s Danny Sullivan argues that BERT leaves “nothing to optimize,” there are indeed adjustments you can make to optimize your SEO efforts and better align with BERT’s changes:
1. Target more informational searches
Informational queries, such as “how to train my dog” or “how to make a chocolate cake,” are broad long-tail phrases that generally come first in the sales/marketing funnel—the type of questions consumers most often ask to generate potential options and start narrowing things down to the point of purchase.
By shifting your focus more toward targeting broader informational searches (“how to train my dog”) than navigational (“dog training methods”) and transactional (“buy dog training 101 online”), you can more often meet customers at the beginning of their online journey—and bring each a little closer to conversion.
2. Create more query-specific content
Think hard about the specific questions your customers are asking before they take that next step. Tracing the journey back to those queries consumers typically start with helps not only set your marketing agenda, but also to design blogs, social posts and videos that provide the answers they’re seeking and improve your brand positioning in SERPs.
Delivering content focused on what the user actually wants to learn rather than trying to cover too much ground may ultimately lead to lower site traffic, but the visits you’ll get will be higher quality. Bounce rates will go down and conversion rates will go up, as you’ve rewarded searchers with the information they wanted as soon as they wanted it.
3. Stick to SEO best practices
It may seem silly, but some of the best advice for how to optimize your SEO for BERT is to avoid going overboard. Google rewards quality, informative content much the same as it did before the BERT update, and sticking to SEO best practices throughout the content creation and curation process remains as important as ever to building brand cred, authority and visibility.
Even with BERT forcing a few minor changes to your content marketing strategy, keeping your content relevant, engaging and informative is key to getting in Google’s good graces—something that won’t end any time soon. Make tweaks and keep your eyes open, but avoid the kneejerk reactions that can upend your efforts and cause spendy issues for your marketing team for the foreseeable future.
What does BERT mean for digital marketing?
While Google continues to downplay the impact of BERT on SEO and content marketing, it’s nearly impossible to know exactly how and how much BERT will affect digital marketing teams in the months and years ahead.
As with any search engine update, Google keeps things pretty close to the chest. And though it’s fairly clear as to BERT’s intention (improving search results for the user), no one really knows how far BERT’s reach will go or what the long-term consequences will be for marketers, SEOs and brands—particularly as the competition for SERPS space grows at breakneck pace.
Diligent research and keeping a close eye on the latest SEO trends are good ways to avoid surprises and ensure your digital marketing machine is running at full steam. But staying ahead of Google and the competition often requires a bit more heavy lifting. And that’s where we come in.
At Marca Global, we provide the outsourced marketing strategies and solutions you need to build a stronger online brand—and to ensure you’re ready for anything Google throws your way.
To schedule a consultation, call Marca Global today at (720) 378-5024.
By: Brandon Purdum
Author Bio: Brandon manages the content team at Marca Global. His focus is writing informative, engaging content that helps brands improve outreach and achieve their digital marketing goals. He is an avid reader, aspiring novelist and father of a precocious four-year-old.