We can’t talk about improving your SEO results without focusing on one crucial element of every content optimization strategy - entities. If you’re uncertain what this term actually means, now is the right time to learn all about it and find out how it can improve your ranking. After all, you’ve probably noticed that adding keywords to your content is simply not enough anymore, right?
Before we start talking about the difference between the keywords and entities, let’s take a look at the entity as a term and what makes it so appealing for all the content marketers around the world right now.
What Is an Entity?
An entity is a concept or a thing that is unique, well-defined and distinguishable. For instance, an entity can be a person, movie, book, idea, place, company, event or pretty much anything that differentiates it from other concepts. As its definition leads to confusion for many people because they equate it with keywords, it’s crucial to understand the difference before your writing process begins. Actually, understanding the difference between keywords and entities is what will provide you with the content marketing results you need.
Entities are crucial for the user search, so using them in the right way will help you tremendously to build your rankings. That being said, entities go beyond the keywords. You will have to think of them as a concept instead of just a word or a phrase for which you are trying to rank. Your content is basically the grouping of all entities connected in a meaningful way.
Entity or Keyword?
As we’ve already said, entity is a universally understood concept, whereas the keyword shows a more narrowly focused term. That’s why all the keyword research tools are providing you with a bunch of different variations of the term you’re researching.
Example: swimsuits, women swimsuits, black women’s swimsuits, etc.
Entity-based research, on the other hand, covers the topic in a more in-depth way with the list of other related entities, but without the same word within the entities.
Example: board shorts, one piece, bikini, etc.
When comparing these two types of research, keyword and entity-based, you can see that the one based on entities provides better information to create quality content. That content will surely be more useful to the users than the one that is keyword-based.
How Google Sees Entities?
Google released a patent in 2015 called Ranking Search Results Based On Entity Metrics, which discusses four factors vital for the ranking of entities:
- Relatedness: The co-occurrence of entities will determine the search. In other words, if there are two entities frequently referenced, it will impact the search results.
- Notability: Within the patent, Google uses a formula to check how important your entity is. The formula shows that the more value your entity has and the value of its category is low, the notability will be higher.
- Contribution: It is a measure that determines the contribution an entity has to a certain topic.
- Prizes: This factor shows all the prizes the entity has received, and it implies its value through the total number of prizes.
As with keywords, when thinking about entities, you will need to think about it from the user’s perspective. What are the entities your prospects and clients are looking for when they search on Google and other search engines?
After you’ve figured that part, it’s essential to understand how Google processes all of this data. The process looks like this:
- Google determines the relatedness of the entities and assigns values to them.
- Google determines these entities’ notability and assigns the value to each entity.
- Google determines the contribution of entities and assigns value to each of them.
- Google determines prizes related to these entities and assigns them value.
- Google determines the applicable weights each of the entities should have based on the type of query.
- Google determines the final score for each entity.
Word2Vec will turn language into a mathematical computation, which allows Google to identify concepts and map them accordingly the way traditional models simply aren’t able to. It is used to provide word embeddings by using a shallow neural network, and the purpose of it is to group the vectors of similar words together in one vector space. These vectors demonstrate numerical representations of the context of individual words.
Without a human interaction, Word2Vec can highly accurately guess the meaning of a word based on its past appearances. These guesses can then be used to create a word’s association with other words or classify them by a certain topic. This can serve as the basis of search and sentiment analysis.
Entities Inform Content and Links
As we’ve already explained entity- or topic-based research above, it’s very clear that your content needs to include the semantic words or phrases related to the entities you have chosen for your text. Think about the way your chosen entities inform the content you are creating and how users will search for it. If Google recognizes your content as relevant to that query, your ranking position will be higher.
In much the same way you are thinking about your entities within the content you are creating, you will need to include the links as well. Following the example from above, if your piece of content is about swimwear for women in Australia, you will need to include links from authoritative sites on swimming, women and Australia. Links demonstrate the relationship between pages on the web, and these pages are also entities with their own entities.
We can’t talk about links without mentioning the anchor text. The entity of your anchor text will show the connection through its relationship to the topic, and that entity will also be connected through a directed relationship to that target page and its entity.
Simply adding your entities to your content will not be enough. Treat them as guidelines on what needs to be covered in your piece and make sure you provide your audience with the information for which they are looking. Answering their questions about the topic is the main goal of any quality piece of content.
Google’s BERT Update
BERT, or Bidirectional Encoder Representations from Transformers, is a Natural Language Processing model presented to the public in October 2019 by Google. This method can consider the entire context of a single word, based on the words before and after named entities. As entities are at the core of NLP models such as BERT, not properly implementing them in your content is an oversight you simply can’t allow to happen.
So, what do you need to know about BERT?
- BERT is Bidirectional, which means it uses the content before and after the keyword or phrase to understand the meaning this keyword or phrase has in context.
- BERT uses a Dataset with Pre-Trained Model to understand the context. Then, it uses a methodology from the data and applies it unmonitored.
How to Optimize Your Content for Entities
Before you start exploring all the entities you see on the list, you will need to choose a topic you want to research. Skipping this step might result in creating content that is not relevant to your audience just because you saw it on the entity list. That’s why it’s vital that you always determine the right topic before you dive into the entity-based research.
If we go back to our example, we can say that our topic of research will be best swimwear for kids. How can you see which topics or entities Google considers related to this topic?
There are a few ways you will be able to better understand what entities Google would suggest to you:
- Wikipedia (use for related keywords)
- Google Image Search (use for related keywords)
- People Also Ask (use for possible headings)
- Google’s NLP API demo (use for checking your article for relevance). To analyze the competition, use this tool. When using the tool, you will see results like the ones below.
In this image, you can see that the entities with the highest salience are brands like “Olga Valentine Swimwear” and “Chickadee Kids Swimwear,” as well as “swimsuits” and an attribute of swimwear “UPF50+.” These would definitely be entities you should include and optimise for in the article about swimwear.
Building Context and Relationships
Don’t forget to build context and relationships through your content. The text you will create should build the relationship between the keyword phrase you have chosen and the entities you have identified in your research.
Go back to Wikipedia and Google Search Images to see which keywords will help you establish that relationship. Explore People Also Ask and see what inquiries people have about your target keyphrase. All of this information will help you write the content that converts!
Using SEO Scout to Make the Entity SEO Easier
So, how can you start using AB Ranking to make the most of entities? Luckily, the process is very simple. Start by inserting your target keyphrase in the search bar under Topic Research. If our target keyphrase is “G Cup Swimwear,” these are the results we will get:
Although all these metrics are relevant for your content, we’ll just focus on entities here. On the right side of the results, you will see the section called Recommended Entities. This list will show you all recommended entities in the top 30 Google results. As they are organised by relevance, make sure you include entities from the top in your article.
If you’re looking into creating several pieces, you can check all the entities and group them into different categories. For instance, from the entities in the image above, you can have three different topics:
- How to Shop for the Perfect Swimwear - swimsuits, shopping, underwire, returns
- Swimwear Trends for 2020 - clothing, styles, collection, printed
- Bikini or Tankinis: How to Choose - bikinis tankinis, sports
Once you’ve chosen your related entities and are ready to start crafting your piece, keep in mind that these entities should connect your entire article. Adding too many entities will result in a low-quality content, not just for Google, but also for your audience.
By using entities in your content, you will provide your prospects and clients with in-depth information. Building long-lasting relationships with your customers is a priority if you know that loyalty is one of the deciding factors when purchasing.
By using entity-based research, your content will cover the topic nicely for your audience and make it easier for Google and other search engines, which use Natural Language Processing to understand and give that content a higher ranking position.