Understanding entities and entity salience is important for anyone working in digital marketing and search engine marketing, especially content strategists and copywriters. Google and other search engines increasingly rely on natural language processing (NLP) to understand language. Identifying entities and measuring entity salience is one of the ways these search engines understand online content.
What‘s the Difference Between Keywords and Entities in SEO?
People trained in traditional SEO and content marketing may be confused about the difference between keywords and entities. It’s important to understand the distinction. Keywords and entities are not necessarily the same thing. A few entities in your content may also be keywords, but most of the entities in your content will not be.
What are Entities?
An entity is any identifiable thing, and all nouns and noun phrases in your content will be counted as entities. Most pieces of content will have at least one hundred entities. Entities will include:
- all names and proper nouns
- locations, including countries, cities, and specific addresses
- dates, months, and years
- abstract concepts
- all common nouns
- noun phrases (for example, “moving company” or “Apple store”)
Natural language processing algorithms used by search engines can also break down syntax and use context to tie multiple words to the same entity. For example, it can understand that the word “she” refers to “Mary” in the sentence, “Mary enjoys her work, and she is good at it.”
What are Keywords?
A keyword is a focused term that should describe the topic of the content. Keyword research and keyword targeting used to be one of the most important aspects of SEO — and they’re still important! But search engines don’t work the same way they used to. Strategies that focus on making sure a certain keyword appears X amount of times no longer work.
Google and other search engines are now able to understand the context of a web page beyond any single keyword. One of the ways it does that is by identifying ALL of the entities in the content. They then use NLP algorithms to identify the overall salience of the content’s entities.
What is Entity Salience?
Entity salience is a measurement of how salient — or important — a specific entity is in the context of any piece of content.
Let’s break that down. We know what an entity is — but what does salience mean? For natural language processing, salience refers to the importance or prominence of an entity in the content. The salience of an entity in a text is how much the text is about that entity. For example, in an article about how to bake chocolate chip cookies, you would expect “chocolate chip cookies” to be the most salient entity.
NLP algorithms scan content and apply a salience score to each entity it identifies. The salience score is the NLP’s prediction of what a human reader would consider to be the most important topic(s) of the content.
If you run content through Google’s Natural Language API demo, it will give each entity in the text a score between 0 and 1. Entities with a salience score closer to 0 are probably not very important to the content. Entities with a salience score closer to 1 are likely more important.
For example, an entity with a salience score of 0.8 is likely VERY important to the content. In an article about chocolate chip cookies, you would likely expect the entity “chocolate chip cookies” to be above 0.2. Note that the more entities in the content, the more diluted the salience score may be for each entity.
You always want the main topics of your content to be represented in its most salient entities. For example, the most salient entities of this article are “entity salience”, “SEO”, “NLP”, and “Google.”
Why is Entity Salience Important for Content Marketing?
Google and other search engines no longer focus only on individual keywords to rank their pages. They are now using advanced natural language processing (NLP) algorithms to understand online content more completely. NLP will only become more advanced in the future.
While the NLP utilized by Google Search is certainly more advanced than the publicly available Google Natural Language API, content marketers can still utilize this resource to see how NLP breaks down the entity salience of their content. This should give you an idea of how Google and other search engines are understanding it as well.
The great news about optimizing for entity salience is that it doesn’t mean you have to write for machines. Writing to achieve higher entity salience also helps make content more clear and crisp for human audiences!
So What Now?
Understanding entity salience and being aware of how search engines use natural language processing to understand content is necessary for anyone writing online content or working in search engine marketing. But what can you do with that knowledge?
If you’re wondering how you can use this to increase the quality of your content, follow my blog! Next week I’ll be posting on how to write salient content to increase your Google Search rankings. We’ll also go over how Google measures entity salience and how you can write to put Google’s focus on the entities you want.