Are You Down With NLP? How Natural Language Search Impacts SEO
Learn how natural language processing (NLP) is changing the face of SEO and what you can do to stay ahead of the curve.
If you've ever tried to solve a Rubik's Cube, you'd know two things. First, they break really easily if you fling them at the floor in frustration. Second, there's a rhyme and reason to solving it (and it'll save your floor from scuff marks).
Natural language processing, or NLP, is a field of computer science and linguistics that deals with the interactions between computers and human languages. This involves creating algorithms that can understand human language and extract information from it. When you boil it down, it's what lets you ask your phone a question and get an answer, or search the internet for something and get auto-complete results.
NLP is complex (dealing with the complexities of human language is no small feat), but the potential rewards are great. After all, if we can get computers to understand us, that opens up all sorts of possibilities—from better search engines to more intelligent chatbots.
And, today, if you want to do SEO well, you'll need to have some basic understanding of how NLP works and how to reverse engineer it to create content worthy of Page One.
NLP's impact on SEO is a recent development. Let's explore how the landscape has changed.
What's Google got to do with it
Unless you've been under a rock since the turn of the millennium, you'll know that Google's always got something to do with content marketing's ongoing evolution.
Google's Hummingbird algorithm in 2013 was a turning point for SEO and modern content marketing. Up until then, Google had been using keyword-based algorithms to match user queries with webpages. So if you wanted to find out how to make an origami swan, you would type in "origami swan" and Google would give you a list of websites that had those two keywords in them.
This worked well enough, but it had its limitations. Keywords are an imperfect way of understanding user intent—after all, people don't always use the same keywords to mean the same thing. "Swan," for example, could refer to the bird, the ballet, or the shape. And "origami" could mean the art form, the paper, or the toy.
Keywords are also a poor way of understanding the context of a query. Consider this search: "How do air conditioners work?" If you were to type in those keywords in some different order (e.g. "Do air conditioners work?" or "air conditioners") you would get very different results. But by typing them together, it's clear that the user is looking for a specific answer.
To get better at extracting the context of a query, Google unleashed its BERT update in 2018. BERT (which stands for Bidirectional Encoder Representations from Transformers) is a neural network that's been trained on a large amount of text. This allows it to understand the context of words in a sentence, regardless of whether the words come before or after each other.
BERT is designed to better understand the context of a search query, and the update is intended to make Google's search results more relevant and helpful for users. For example, a search for "2019 brazil traveler to usa need a visa" will return results based on the user's intent, and not just their keyword matches.
In addition to improving the accuracy of search results, Google says that BERT should also result in more relevant ads and better ranking of featured snippets. As we've seen in the past, search results that include snippets have a higher click-through rate, which could translate into improved ad performance and revenue.
Why is BERT important? Because it was built as a language model for NLP.
How NLP works
Now to the nitty gritty. NLP algorithms have to break down text to know what to do with it.
Parsing is when these algorithms split text to derive meaning from it. To understand the nuances of text, these algorithms rely on a process of sentence segmentation and tokenization, which breaks the text down into smaller pieces that can be more easily processed. Once the text has been tokenized, it becomes easier to glean the larger meaning behind the original text.
Lemmatization is a process of reducing a word to its root form. For example, the words “walking”, “walked” and “walks” all have the same stem: “walk”. This is important because it means that NLP algorithms can more easily identify the relationships between words.
Part-of-speech tagging is the process of assigning a word to a particular part of speech, such as a noun or verb. This is important because it can help to identify the meaning of a word based on its context.
Named entity recognition is the process of identifying proper nouns within text. This is important because it can help to identify people, places, organizations, and other proper nouns.
Sentiment analysis is the process of identifying the sentiment of a text, such as whether it is positive or negative. This is important because it can help to identify the overall tone of a piece of text.
Topic modeling is the process of identifying the topics that are present in a text. This is important because it can help to identify the main themes and ideas within a piece of text.
Stop words removal is when the algorithm will automatically remove stop words in your text. Stop Words are common words that do not provide additional information or meaning to a passage of text such as "and," "the," and "a." This can help to make the NLP process more efficient by removing unnecessary words and allowing algorithms to focus on the more important aspects of the text.
Using NLP to improve your SEO
All of those NLP particulars means marketers have to change how they look at effective SEO in the modern era. There are a plethora of ways to incorporate NLP thinking in your SEO strategy, but two are especially important:
- Topic modeling. In recent years, the majority of SEO experts' attention has shifted from keyword targeting to cluster targeting. This semantic SEO strategy can help identify the main topics and concepts on a website and make it easier for search engines to match relevant queries with the appropriate content. Take it one step further and identify NLP entities that Google values (related to your content's topic).
- Content Optimization. NLP can also be used to optimize your content for better search engine visibility. This includes things like using the right entities and phrases, as well as ensuring that your content is well-structured and easy to read. Using these little advantages can bump up the overall quality of your content and improve your chances of ranking higher in search results.
Optimizing your content for today... and tomorrow
The best way to win in the modern SEO war? Keep creating great content. Don't say what everybody else is already saying, and don't over-optimize with fluffy and superfluous keywords.
And don't just focus on your written content. Google's giving its users video suggestions wherever relevant, and voice assistants have apps that share relevant bits of audio at a moment's notice. Optimizing for today's search user intent and the modern web user means considering all the channels your content should ultimately thrive on.
NLP is here to stay. There's no fighting it. By understanding how it works and incorporating it into your content creation process, you can create better and more meaningful content. And if you need the strategic support to get you there, there's always fractional marketing partners that are a click or call away.