Google recently rolled out what some are calling its biggest algorithm change since RankBrain. BERT (“Bidrectional Encoder Representations from Transformers”) is a neural network-based technique using natural language processing.
The new algorithm update is expected to impact one in every 10 searches, and help Google better understand nuances or context within a search query. In its blog post about BERT, Google comments that despite having insignificance in keyword searches, prepositions such as “for” and “to” carry significant meaning within conversational queries.
For example, in the search, “2019 brazil traveler to usa need a visa,” the word “to” is important to understanding the meaning of the query. Thanks to BERT’s rollout, Google understands that this query is related to a Brazilian traveling to the U.S. and not the other way around.
“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,” said Pandu Nayak, Google Fellow and Vice President of Search.
So what does this all mean for marketers? Google has already commented that BERT isn’t something to optimize for. However, it does carry some implications for businesses.
I spoke with The Search Agency’s creative search engine optimization team to learn how brands should interpret the latest BERT update. Here are four takeaways from the latest announcement.
1. BERT is ultimately about voice search
Google understands that its search product isn’t perfect, but it’s getting close. As Nayak confesses at the beginning of his blog post, “We see billions of searches every day, and 15 percent of those queries are ones we haven’t seen before.”
BERT helps Google close the gap on that 15 percent of new searches by adjusting its algorithm to the way people speak. Naturally, Google’s emphasis on learning how humans speak further suggests that the search engine is hedging its bets on voice search becoming more mainstream.
“This update is clearly optimized for voice search. Since Google is continuing its on-going effort to understand voice search, brands need to do so, as well.”
–Gregory Sidor (Director, Earned Media)
This means brands should think less about optimizing for keywords and more about which phrases or questions your company is relevant for.
2. Brands need to write content that answers questions
Speaking of voice search, Google has made it clear that it believes questions are more natural to users than keywords. If Google is hedging on the success of voice search, then optimizing towards questions makes a lot of sense.
Users will want to receive information in the same conversational manner that they’d receive it in talking to a friend. Google admits that its search product still falls short of this goal:
“While we’ve continued to improve our language understanding capabilities over the years, we sometimes still don’t quite get it right, particularly with complex or conversational queries,” Nayak said. “In fact, that’s one of the reasons why people often use ‘keyword-ese,’ typing strings of words that they think we’ll understand, but aren’t actually how they’d naturally ask a question.”
Ultimately, this means brands need to focus their content on answering questions that their customers care about. The more questions that brands can answer, the better return they’ll receive on their SEO efforts.
“We need to help Google understand our answers better by writing good content and structuring it to target featured snippets and answer boxes.”
–Aryn Kennedy (Senior Director, Earned Media and Creative Services)
Answering questions is all about winning the featured snippet. However, content alone won’t make that happen. Tactics such as structured data markup help indicate to Google which questions a specific piece of content is designed to answer.
In addition to driving terrific SEO value, answering questions provides an opportunity for brands to achieve better awareness and become thought leaders within their industry.
But which questions are most valuable? Try asking customer service.
3. BERT is a win for customer service
Prior to Google’s existence, prospective customers either relied on sales or customer service to assist in the research process. Today, customers are doing all their research online and getting their information from third parties such as articles or review sites.
While third party information is valuable, it can’t provide the deep product knowledge that only your customer service team owns. Therefore, brands should view the BERT update as a nod towards leveraging customer service as a method for ranking better on search.
“Brands should ask their customer service reps what they get asked the most and write answers to them on the website. It would not only benefit the company from a search perspective, but also save the customer service team a lot of time and money.”
–Amanda Haxton (Account Manager, Earned Media)
That information can then be distilled into valuable content, which will help prospects better qualify themselves before reaching out.
4. BERT appears to benefit brand-related searches
This update seems to ultimately help brands own their terms.
While this may not seem like a big deal at first glance, it’s an encouraging move for industries like retail where smart aggregators have discovered ways to outrank brands for their own terms. BERT appears to be putting a stop to this.
“Since BERT went live, we saw a big boost in featured snippets for brand-focused searches among our clients. This ultimately helps brands show up when a user is searching for specific information.”
-Gregory Sidor (Director, Earned Media)
The BERT algorithm update is an important step towards helping Google understand natural language queries. This means that now, more than ever, brands need content that’s well organized and directly benefits its readers.
Ready to discuss how the BERT update will impact your business? Contact us today for a free consultation.