Entity SEO 101
Also known as entity optimization, entity SEO is content optimization using entities instead of (or to complement) keywords.
This is a type of optimization that is taking a bigger and bigger role in semantic SEO.
We’ll see in this guide what entity SEO is all about, starting with the most foundational of questions: what is an entity?
What is an Entity?
According to Google, “an entity is a thing or concept that is singular, unique, well-defined and distinguishable.”
That means that an entity is:
- One (singular)
- Such that there’s not another one like it (unique)
- Something that has certain qualities (well-defined)
- Separate or different from other entities (distinguishable)
and it might be anything from a person to an abstract concept, and it’s represented by a noun in a language.
That’s quite a lot but still clear enough, isn’t?
An entity is different from a keyword because, while an entity defines something in its wholeness (i.e. a thing, a concept), a keyword is a description or the verbal expression of that thing in a given language (e.g. “capital of Italy” is a description of “Rome” in the English language).
If you’re thinking about “web calendars” you may also think Asana and Google Calendar. That’s what Google’s algorithm may think about, too. They can be used as keywords, but in fact they are entities.
So, from everything we’ve seen so far, we can say that entities can be:
- Consumer Good
- Work of Art
Entities and Relationships
Entities can be linked to each other in a graph (with entities as the nodes and edges as the relationship).
Entities are the foundation of Google’s Knowledge Graph and they can also be linked to a Knowledge Base containing entity descriptions and any other information about them.
It’s easier to understand entities if we imagine them as the entries of an encyclopedia or Wikipedia — and these entries are linked to other entries through links or bibliographical notes.
But let’s see it in practice with an example.
In the sentence “buy cheap textbooks to save money in college” there are 3 entities:
linked by relationships.
Below is a graph that shows the 3 entities and the relationships that link them together:
This is easy enough, isn’t it?
And Google Cloud NL API demo tells us exactly the same thing about this sentence:
This leads us to the next question: how are entities used by Google to return search results?
How does Google use Entities?
A 2014 patent titled “Ranking search results based on entity metrics” mentioned relatedness, which depends on co-occurrence, meaning that
the more frequently two entities appear in the same text, the better they are related.
As an example, “best superhero movies” and “iron man” are likely to be mentioned together, leading to results such as the ones below:
Naturally, a lot of factors contribute to ranking, and entities are not a ranking factor in the traditional sense (see later in this post) but they contribute to the shaping of the link graph, and given that Google now uses more machine learning than in the past, focusing on entities as an SEO practice is more than justified.
How to Do Entity SEO?
There are three methods essentially.
Method #1 – Google Images
Take a look at Google Images: searching for “best robot kit” will return a list of entities below the search field, that Google associates with the key phrase:
Here the entities Google Images gives us are:
- arduino robot
- kits for
- humanoid robot
- arduino uno
- educational robot
- jimu robot
- robot toys
- lego mindstorms
If we were to write content on “best robot kit”, we would make sure to include all these entities.
Method #2 – Google’s Natural Language API Demo
You can also use Google’s Natural Language API demo (we introduced it earlier but it’s worth linking again) to find entities in a text.
For example, let’s take this article on resource page link building and input a portion of its text in the demo field and click Analyze:
We found the following entities:
- Resource page
- Link building
- Resource page link building
- website owners
- link request
and others (a total of 16 entities).
We know that not all of these look like entities, some look more like keywords — but you can easily realize that they are entities because they don’t describe something, they are something.
The tool also gives us a significant piece of information about each entity: salience.
Salience is the score that tells us how important a word (or group of words) is a text.
In our example, “Resource Page Link Building” has the most importance in our text.
What term has the highest salience in yours?
You can also use the API demo method to find out how to optimize your content for entities:
input your target keyword or key phrase and then see what entities (and other keywords) the tool suggests.
Finally, you can run a competitive analysis for entities using this demo tool: simply input a competitor’s content (the text from one of their pages) in the analysis field and see what results the tool gives you.
Remember to make sure that’s a competitor appearing on page one for your target query for better data to build your content upon.
Method #3 – Wikipedia
You can search your topic in Wikipedia and see what other links or entities that Wikipedia page has in its text.
That will give you an idea of what other topics you can cover in your content that are related to your target one.
For example, the Wikipedia page for Content Marketing gives us, in the introductory text, a list of entities that are strictly related to the concept of “content marketing”:
- target audience
- white papers
- email newsletters
- case studies
- how-to guides
- question and answer articles
It makes sense because these entities are all part of content marketing. If we had to draw this as a graph, we would have “content marketing” as a central node and then all these entities as other nodes linked to it.
If you don’t want to make this entity analysis yourself, you can run the text of the Wikipedia page you’re researching in Google’s NL API demo and see what entities it returns.
There are a number of other optimizations you can do on your content that use entities:
- Optimize for People Also Ask because that gives you related entities you can cover in your content
- Don’t forget search intent — understand what meaning Google gives to specific queries and give it an entity name (e.g. for the keyword “solving puzzles”, Google gives us results on a variety of puzzles you can solve to hone your problem solving skills — so the associated entity is “puzzles”, and you can cover it in its entirety in your content)
What the Experts Say on Entity SEO
We asked experts on HARO to share their insights on entity SEO with our readership.
Specifically, we asked:
- How do they use entities in your SEO?
- How do they optimize for entities rather than keywords?
Here’s what they told us.
Nikola Roza (CEO and Owner)
I optimized this page on my site for entities, and not just for keywords. Here’s how!
- I scooped up all Google Suggest queries and made them part of my content outline as subheadings and sub-topics.
- I took advantage of PAA boxes that were showing up at the time (now they’re gone) to see what users ask about the topic.
- I used related searches at the bottom of the page, as these are powered by the Knowledge Graph.
- I went to Google Images and found new entities displayed in the suggested searches bar above the first row of images. These are also entities delivered by the Knowledge Graph.
- I went to Google NL API and pasted the text of the first 3 ranking pages. I scooped up all terms with high salience scores.
- I went to YouTube and found new queries which were auto-suggested in the YouTube search bar.
- As a final step, I wrote my article naturally and then came back and tried to include as many entities as I could, in a natural way of course.
I think it worked, because I currently rank on the second page for 20+ keywords, and my page is less than 3 months old.
It’s showing great potential!
Eduard Dziak (Marketing Manager)
Often I am using entities to learn more about semantic meaning that Google associates with the targeted keyword or topic I am going to write about. Understand the particular important key phrases that Google associate with my target topic helps me guide my content and ensure that I provide a comprehensive article with answers that Google and its users are looking for.
Because optimizing only for a keyword often lead to keyword stuffing and doesn’t deliver enough value. Also, I have noticed that the content often leads to fewer indexed keywords if you only optimize for the keyword rather than for semantic and entities.
So to optimize my article for entities I am using keyword research tools along with natural language processing tools from Google and MonkeyLearn Keyword Extractor. These tools point out to me what other keywords are associated with my targeted keyword or topic and thus navigate my content to create SEO optimized article.
Like this I can create articles that I know will rank well and provide the content Google and users are looking for when they are typing my targeted keyword phrase.
Jenna Carson (Marketing Director)
The best way that we’ve started to incorporate entities within our SEO strategy has been to use software that uses extraction of relevant entities for you and identifies Google’s understanding of them – this helps us integrate them within our other SEO techniques and plans.
We do our best to optimize entities by referencing the music entities we’re experts in within a lot of our copywriting and blogs – however, we’ve had the most success in the first quarter of 2021 by stopping the typical keyword-stuffing SEO behavior. We have a list of entities we want to focus on and we now carefully craft content around this, before publishing any articles we look at competitors’ entities’ successes and see how we can tap into this.
Anatolii Ulitovskyi (Digital Marketer, Founder)
Generic SEO is related to finding keywords, optimization them, and providing link building. That’s not the case anymore because of huge competition with around 1.8 billion websites that want to get traffic.
Modern SEO is related to topics [entities], not keywords. Google understands the user intent and ranks content that users want to get. Instead of searching for generic keywords, find topics with a lack of quality content and low competition and jump there.
These topics are not always relevant 100% to your niche but they cover all the stages of a sales funnel and build brand awareness. Most projects fail because they chase high-volume. Over a thousand others do it the same. Search on Google any keyword. There are a million results.
Search systems have a lot of choices. It’s not possible to rank even a tiny percentage of existing content. That’s why I search for topics with outdated content that my competitors ignore.
Zander Buteux (Head of Organic Growth)
At Wilbur Labs, we use a few tools to help us as we generate content at scale — around 40 to 50 pieces per month today, and we anticipate generating 100+ by the end of Q3. Copy.AI, Fraze, and SEMrush are the go-to platforms for us. Copy.AI helps turn a blank page from a keyword into a title, into an outline, and then an intro paragraph. Fraze and SEMrush help turn the outline into not just a body of well-written words, but one that is optimized for the topical entity. Whereby it provides recommendations of questions to answer, related keywords to include, and suggested length of the article based on SERP competition.
This tactic has worked exceptionally well for Joblist, which tackles content generation with a spoke and hub strategy — “hub” being a parent piece that discusses a topical cluster of clusters and the spokes being a deeper dive into the nuanced topics within. Using this strategy of generating strong, entity-focused content that truly helps and guides our target market, our content hub has grown on-site traffic by 1230% Y/Y.
The biggest takeaway here is that keyword chasing will produce momentary growth that will require more hands-on updating over the life of the piece, while responding to entities will generate sustainable growth over the long term and nightmares of algorithm updates ripping away 90% of your traffic can be a thing of the past.
John Pinedo (Founder)
Internally linking entity-related articles helps Google recognize your website as an authority on the entity topic.
If I’m covering an entity topic like landing page software, I’d want to cover as many informational and commercial topics as possible on my site. Google will eventually crawl and index my articles on these entity-related topics, but it’s not until they’re internally linked together that Google starts registering that my website has covered said topic thoroughly.
Soon enough, when someone searches for an entity-related topic, Google can match up to my website as a great fit for the searcher (ranking me higher in the SERP) by cross-checking its knowledge graph with my on-page SEO and site structure efforts.
Are Entities a Ranking Factor?
Not if taken as a standalone entity (pun intended).
Like keywords, entities are not a ranking factor per se, but they can be as something that goes together with content.
Entities are the way content is understood by Google’s natural language processing model and related with other content.
That — content — is a ranking factor, as well as links and RankBrain.
Conclusions on Entity SEO
Entity optimization is the future of SEO. It doesn’t replace keyword research combined with search intent, but it gives a broader meaning to it, for topics are not just keywords — they are entities, which exist regardless of language.
We recommend using the methods we outlined in this guide to do your entity SEO and come up with entities to use in your content.
For further help and guidance, take a look at our On-Page SEO services.
Here’s to your success!