Call Us! 888-340-6731

What is the BERT Update and What Does it Mean for You?

posted by Michael Epps Utley Michael Epps Utley
What is the Google BERT update

Google algorithm updates, in theory, make Google easier to use. Each change is designed to interpret search data more intelligently, so users can find exactly what they’re looking for sooner, ideally without having to navigate off page one.

But Google changes their tactics fairly often based on search data. So it’s important for you to be aware of the latest Google algorithm updates, and to have a clear understanding of what it means for your SEO, your content marketing, and your overall marketing strategy.

To that end, we’re taking a look at Google’s BERT algorithm. We’ll demystify this update and give you practical ideas for how to use it to your advantage in your marketing strategy.

WHAT TO KNOW ABOUT GOOGLE’S BERT UPDATE

What Is It?

The BERT update is one of Google’s latest algorithm updates. BERT, which stands for Bidirectional Encoder Representations from Transformers, is a deep learning algorithm related to natural language processing. It helps a machine, such as a search engine bot, to understand what words in a longer search term mean by using nearby words as context.

When interpreting what someone types or speaks into the search bar, BERT handles tasks like:

  • Entity Recognition: BERT does better than previous algorithms at understanding who’s who (or what’s what) in a search query.
  • Part of Speech Tagging: BERT recognizes when words are subjects, objects, actions, and so on.
  • Question-Answering: BERT does its best to answer specific questions using context clues.

Why Is It Important?

Weighing the importance of Google algorithm changes helps you prioritize content, SEO and search advertising improvements.

BERT is a critical update because its changes could affect 10% of all search queries. It represents an improvement for the Google user, since Google has become more likely to return relevant results based on context.

That means, for example, that you could search for “Louise Penny books in order” to see the sequential order of books in that author’s famous series.

Screen Shot 2020 11 10 at 12 37 52 PM

Because of BERT, Google didn’t throw out the words “in order”, but used them as context for the more specific keywords in your search. The result? You get exactly what you’re looking for, without having to worry that you’re starting a series in the middle.

What Does it Mean for SEO?

BERT has several implications for search engine optimization:

  • Keyword Stuffing Won’t Work: No one is saying you have to remove keywords, or that you can’t add more detailed keywords to your website. That’s always a good idea if you have data. But “stuffing” content with keywords by putting lots of them in where they aren’t a natural fit won’t help you do well in the age of BERT. That’s because no one’s natural language repeats the same keywords over and over in a short amount of time.
  • Long-Tail Keywords Are Good: Long-tail keywords are longer groups of words and phrases. Instead of “industrial painter Boston”, a longtail keyword might look like “hire industrial painting contractor in Boston”. Long tail keywords can be questions, phrases, or sentences, and typically include multiple related keywords.
  • Understanding Audience Keywords Is Crucial: Often, the words you use to describe your products and services are different from the terms they use to search for you. Completing and regularly updating keyword research to include terms your audience is using to find you will mean you have their “natural language”, which BERT is also using to rank results, on your side.

What Does it Mean for Content Marketing?

Optimizing for BERT and other Google algorithm updates means making sure your content marketing is tied to SEO, and vice versa. BERT should be influencing your content in the following ways:

  • Long-Form Content: Longer-form content makes long-form (or long tail) keywords easier to include. If you are already creating new content regularly, aim for a 500 word minimum for new pages. You can do a deep-dive and make content longer, but keep it relevant. This new update makes it clearer than ever that relevance matters.
  • Accurate Content: Since BERT allows for greater specificity in search results, you want to make sure your content is accurate. Incorrect or outdated content isn’t going to get you more conversions.
  • Unique Content: If BERT is turning up search results that are more aligned with specific questions we ask, our content marketing needs to include unique content. Take a service area you focus on and explore it from a variety of angles: what is or isn’t true about your services? What are common mistakes people make when hiring someone in your field or industry? Is there anything you can tell your target audience to help them save time and money? Building your content to answer specific questions isn’t just good for the BERT update, it demonstrates the authority and expertise you need to build trust with your customer.

How Does it Change SERPs?

By prioritizing content relevance, BERT is poised to change Search Engine Result Pages (SERPs), and how you make it there, in a few ways:

  • Keyword density will become less important than keyword specificity and relevance.
  • Featured snippets will become a regular part of SERPs, with the specific answer to a user’s query bolded on pages where it appears.
  • Since the technology behind BERT is open-source, you can expect this trend toward natural language to become relevant across search engines and databases in years to come.

Google wants its results to be as fine-tuned to our search queries as possible. That means that, beyond BERT, we can expect more and more updates to recognize exactly what our customers are asking for. Make sure your content marketing, your SEO, and your paid search strategy are working together to deliver that. If you have questions along the way, GoEpps is here to help.

We Are the Digital Marketing Pros

Work with a great team of passionate, experienced professionals.