The Discovery Phase: What is NLP, and why does it matter?

If you've ever wondered how Google classifies high-quality content from so-so content, this lesson will shed some light on the subject!

Transcript & mp3 file 👇

Oh hello there! Danni from Surfer here! Welcome back, happy to see your face! So, this lesson is going to talk about NLP and BERT. If you don't know what that means, or what it stands for, don't worry. Just keep watching!

Let's dive in, shall we!?

Here’s what we know affects rank on Google: backlinks, site performance, word count, and domain authority. Easy-peasy, no?

Another crucial ranking factor that Google considers is content, and that’s where things can get slightly more complex.

SEOs have been trying to make content as relevant as possible for a long time. We all know we need to include words and phrases that Google associates with our niche. This is why old techniques like keyword stuffing (which means putting as many variations of the main keyword with an exact match as possible), manipulative anchors for internal linking, or creating several pages for every version of the main keyword exist.

Because, in theory, a page that uses the main keyword multiple times, over and over, is clearly more relevant to its niche than the page that only uses this keyword once or twice, right? Well, not anymore. Google’s algorithms are getting smarter and smarter every day.

Google tries to understand the content on the webpages the way users do. Real people like you and me.

They use both simple and advanced algorithms to do so. This is why spammy content won’t get you anywhere anymore, no matter how many times you force the keyword into the text. This is also why it’s not enough to just put your keywords in your headings and subheadings and Boom! Call it a day!

Google wants QUALITY. So sorry.

The great part is that there’s another algorithm we’re SURE Google uses to analyze the content… and we can use it as well! It’s called BERT.

The BERT algorithm is based on NLP, Natural Language Processing, which means it’s backed by machine learning and not only a simple equation, and that it keeps learning new things all the time.

With BERT, Google is now better at understanding the context of each sentence in the article. It no longer treats words as separate instances but is able to grasp how they fit together.

It helps Google comprehend long-tail keywords, and niche topics that could otherwise be misinterpreted.

Wow, so a lot of information so far. Just want to do a quick pause to make sure that you're understanding and that things are clear!

When we're talking about BERT and keyword density, there are three topics that we're going to dissect even further: entities, category, and salience. Ready?

The entity is a word or phrase that represents an object which can be identified, classified, and categorized such as persons, goods, events, numbers, organizations, etc.

For Google, “Samsung” is no longer a random word; it knows it’s about phones. So, if you randomly place the word Android in a text about baking vegan cookies, there’s no way it will rank in a SERP where people search for the best Android phones.

Next, we have categories! Thanks to NLP, Google can categorize content based on its understanding. For example, Internet & Telecom/Mobile & Wireless. So Samsung would fall into that category, but unfortunately, vegan cookies would not.

I really love vegan cookies… ok, but I digress. Let's talk about salience!

The salience in NLP represents the entity's importance in the text. This ranges from 0.0 to 1.0.

The higher the salience value, the more important and relevant the entity is for the subject of the page.

For instance, the word “Galaxy” may be more important than “iOS” when talking about Samsung.

So I suppose the big question here is: how do you factor in all these components without losing your head at the same time?

Tools! You need tools friends!

You can use any SEO tool that analyzes content, like Semrush or Marketmuse. But let’s be real, Surfer is leading the charge! It’s a total game-changer!

At Surfer, we have our own algorithm that incorporates semantic and correlation analysis, website rendering, and machine learning (which is actually more thorough than Google’s BERT methodology and their NLP API). So you can be sure our recommendations will always be super informed and data-driven!

If you use our Content Editor, you can receive accurate guidelines on prominent phrases to use only by setting up a search for your keyword, and all in real-time. You’ll be told which keywords to include, how often to include them, which should be written as headers, and you’ll be notified if your keyword density gets too high so that you can focus your attention on the important part: writing quality content your audience will love. What is super cool is that we put these keywords directly in the Outline Builder. See those highlighted terms? They are the most prominent terms!

I'm not a mind reader, but I can almost feel that your shoulders dropped at least an inch after hearing this information. I hope you're feeling more relaxed and more confident. I think that's enough for now, at least for this lesson.

In the next lesson we're going to talk about why it's so important to meet your customer where they are on their buyer's journey.

In the meantime, Happy Surfing! I'll see you in the next lesson.

Last updated