Understanding Entity SEO: What it is and how to Optimize for It [2024]

(Need help with your site’s SEO? Reach out today to ask about our SEO Consultation services!)

Entities provide structure to the otherwise unstructured data on the web, are you currently optimizing for entity SEO? You would be surprised by how important this is and how much opportunity is being left on the table.

Generative AI (products like Chat-GPT and Midjourney) are disrupting the search landscape due the influx of AI content and regurgitated information. This is creating a ton of noise in the industry leading to an increase in demand by Google, Bing and other search engine companies to tether their ranking algorithms to something solid. (Well, metaphorically speaking anyway)

That something, is “things”. Another word for these “things” are entities.

To rank in today’s ecosystem, you have to first understand what entities are and how they came about.

Entities are the foundation of machine learning and companies like Google and Bing have been using them long before Chat-GPT to evolve their multi-billion dollar search algorithms. We are witnessing a seismic shift in our understanding and application of strategies. SEO professionals who stay ahead of the curve by understanding entities, and entity SEO, will have a clear advantage in the digital space if they grasp the foundational concepts in this article.

Looking back to look ahead: a brief history of Google's entity search system

Google's transformation to an entity-based search system initiated with the acquisition of Metaweb Technologies in 2010, the company behind Freebase, an open-source database of over 12 million structured data items. This step marked Google's shift from keyword-based search to an entity-based model, encapsulated by the catchphrase "things, not strings."

Subsequently, the merger of Freebase into Wikidata, a structured data project by the Wikimedia Foundation, further enriched Google's entity database, leading to more contextual and accurate search results.

Schema markup emerged as an indispensable tool in this evolution. As a standardized format for classifying webpage content, Schema allows webmasters to explicitly convey to search engines what their site's content signifies, not merely what it says. This approach helps search engines like Google better understand and index web content, significantly impacting SEO performance.

Thus, Google's transition to an entity-based search system began reshaping SEO, demanding a move from traditional keyword strategies to more sophisticated, entity-based approaches concentrating on embeddings and (in the earlier days) spoon feeding Google like baby to understand what your site was about using schema markup.

The evolution of Google's ranking algorithms

As time went on, Google built more complex “baby algorithms” that included more and more specific vectors to rank against. Here's a closer look at just a few of these vectors:

Word Vectors: These represent the semantic meaning of individual words. For example, in the vector space, "dog" and "puppy" would be closer together than "dog" and "tree."

Website Context Vectors: These vectors capture the overall context of a website or a web page, including its theme, topics, and the kind of audience it caters to. They allow Google to understand the overarching message or purpose of a website.

Author Vectors: These vectors signify the credibility and authority of the author, including their expertise and influence in a particular domain. An authoritative author writing on their topic of expertise would likely boost the page's ranking.

While we don’t know all of the details, we can begin to understand more and more about vectors and entities by referencing schema libraries (like at schema.org) to have a clearer picture of which parts of our digital presence can be optimized for entity SEO. (eg. A cooking blog should be using JSON-LD schema markup for their content around recipes and tagging their website property as “same as” if they have a social profile or are found on Wikipedia.)

Decoding 'embeddings'

The concept of 'embeddings' might sound complex, but it forms the backbone of Google's innovations in understanding human language and representing it in a manner that's machine-readable. Also, generative AI uses the foundation of embeddings to understand input and produce output based on these same principles.

There are multiple types of embeddings, but let’s keep it simple. let’s use a word embedding, which represents words as numerical vectors.

Here's an example using word embeddings:

Word: "cat"

Word Embedding Vector: [0.2, 0.4, -0.1, 0.8]

In this example, the word "cat" is transformed into a vector representation within a multi-dimensional space. The vector [0.2, 0.4, -0.1, 0.8] captures the semantic meaning and context of the word "cat" based on its relationship with other words in the training data.

Similarly, other words like "dog" or "tree" would have their respective word embedding vectors, which can exhibit different values based on their semantic similarities or differences.

Let’s bulletize this next part to follow the concepts in an easy to understand way:

Understanding Embeddings in the Context of SEO:

Embeddings are mathematical representations of language that convert words and abstract concepts into vectors within a multi-dimensional space. (Important for the next section of this article!)

Embeddings play a crucial role in Google's innovations in understanding human language and enabling machine-readable representations.

These representations are used in various natural language processing (NLP) and machine learning algorithms.

Embeddings help search engines like Google comprehend the meaning, context, and relationships between words and concepts.

In the context of SEO, embeddings are relevant for optimizing content and improving search engine understanding.

By utilizing embeddings, search engines can better interpret and match user queries with relevant content.

Explaining Embeddings:

Embeddings transform language into numerical vectors.

Each vector represents a word or concept and its semantic meaning.

Words or concepts with similar meanings have vectors that are closer together within the vector space.

This allows algorithms to understand the contextual relationships between words and concepts.

Why Embeddings Matter in SEO:

Embeddings enable search engines to better understand content and user queries.

By leveraging embeddings, SEO practitioners can optimize content to align with search engine understanding.

Optimizing content for embeddings helps improve relevance, semantic understanding, and matching with user intent.

Schema markup and structured data can also contribute to enhancing the understanding of entities and embeddings by search engines.

Cosine similarity: measuring semantic closeness

Once we've represented words or concepts as vectors, how do we quantify their semantic similarity? This is where 'cosine similarity' comes in. (And trust me, I’m definitely speaking outside of my comfort zone here, but hang on tight…)

Cosine similarity is a measure used to determine how similar two vectors (or items) are within this multi-dimensional space. Instead of simply comparing words based on their appearance (as in traditional keyword-based SEO), cosine similarity allows us to measure the 'closeness' of words or concepts in terms of their meaning or context.

A striking illustration of cosine similarity is demonstrated in the Word2Vec model, where a vector operation resulted in a semantic relationship (source). By subtracting the vector for "man" from "king" and adding the vector for "woman," the resulting vector was closest to "queen." This underpins the ability of vector space models to understand and replicate human-like word associations.

Let's break it down:

Vector Space: Picture an infinite multi-dimensional space where each point, or vector, represents a specific word or concept. The idea is that semantically similar words are positioned close to each other, while different words are further apart.

Semantic Relationships: The placement of vectors isn't arbitrary; it's based on their semantic meaning and contextual relationship with other vectors. In this manner, embeddings can encapsulate complex relationships between words and concepts.

Machine Readability: By converting language into a numerical form, embeddings enable computers to 'understand' and process human language, revolutionizing fields like machine translation, sentiment analysis, and, of course, SEO.

Entity SEO: the future of search optimization

Entity SEO, where search engine optimization is tailored towards entities rather than keywords, is widely considered the future of SEO. (source) But to win, you’ve gotta get in-bed with the embed. (Oh man, this rhyme though… I’m dead.)

Here’s a few factors to keep in mind when optimizing for the concept of your target entity:

Content Optimization:

Use word embeddings to analyze the semantic relationships between words in your content.

Identify related keywords or phrases that have similar vector representations to your target keywords.

Incorporate these related keywords naturally in your content to improve its relevance and comprehensiveness.

User Intent Alignment:

Utilize word embeddings to understand the underlying intent behind user queries.

Identify the intent clusters or patterns within the vectors that correspond to specific user needs or search intent.

Optimize your content to align with these intent clusters by providing the most relevant and useful information.

Keyword Research:

Leverage word embeddings to expand your keyword research.

Identify synonyms, related terms, or contextually relevant keywords based on their vector proximity to your target keywords.

Incorporate these additional keywords into your content strategy to capture a broader range of search queries and increase visibility.

Natural Language Generation:

Utilize generative AI models that rely on embeddings to generate high-quality and relevant content.

Train or fine-tune these models on relevant datasets to create SEO-optimized content at scale.

Generate content variations that align with different intent clusters or target specific user needs.

Entity Optimization:

Apply entity embeddings to enhance entity optimization in your content.

Use entity embeddings to identify related entities or attributes that have close vector representations to your target entities.

Optimize your content to include comprehensive information about these related entities, thereby increasing its relevance and depth.

Here's a few levers to pull when optimizing for entity SEO:

1. Comprehensive Content:

Create in-depth and authoritative content about the entity.

Provide detailed information, history, attributes, and relevant context.

Example: If targeting "Tesla" (entity), create comprehensive articles about its electric vehicles, sustainable practices, and technological innovations.

2. Entity-Specific Keywords:

Incorporate specific entity-related keywords in your content.

Example: If optimizing for "apple" (entity), include keywords like "apple products," "apple company," or "apple news" in relevant sections of your content.

3. Structured Data Markup:

Utilize structured data markup (e.g., Schema.org) to help search engines understand entity-related information.

Include structured data elements such as entity name, description, type, attributes, and relationships.

Example: Implement structured data markup for a product entity, including its name, description, price, reviews, and availability.

4. Entity Associations:

Establish connections and associations between entities in your content.

Link related entities and provide context for their relationships.

Example: If discussing "coffee" (entity), mention its associations with "coffee beans," "brewing methods," "coffee shops," or "coffee culture" to create a comprehensive entity network.

5. Knowledge Base Contributions:

Contribute to reputable knowledge bases or collaborate with authoritative sources.

Add accurate and up-to-date information about the entity.

Example: Contribute to Wikipedia by expanding an existing page or creating a new page dedicated to a specific entity.

6. Entity-centric Link Building:

Build high-quality backlinks from authoritative websites related to the entity.

Focus on obtaining links from reputable sources that discuss or reference the entity.

Example: Seek links from industry publications, expert blogs, or news sites covering topics related to the entity.

7. Social Signals and Engagement:

Foster social signals and engagement around the entity.

Encourage social sharing, comments, and discussions related to the entity.

Example: Promote your entity-focused content on social media platforms and encourage users to share their experiences or opinions.

8. Entity-specific Landing Pages:

Create dedicated landing pages for important entities.

Craft unique, informative content tailored to each entity.

Example: Develop a landing page specifically focused on "New York City" (entity), highlighting attractions, history, and travel tips.

9. Local Entity Optimization:

Optimize for entities with local relevance, such as businesses or landmarks.

Include localized keywords, address, phone number, and operating hours.

Example: Optimize a restaurant's website for local entity SEO by incorporating location-specific keywords and providing accurate business information.

10. Entity Cross-Referencing:

Cross-reference entities within your content.

Link to related entities or provide references to enhance entity connections.

Example: If discussing "Star Wars" (entity), cross-reference characters, planets, or iconic quotes to enrich the entity associations.

11. Entity Image and Multimedia Optimization:

Optimize images and multimedia content related to entities.

Use descriptive filenames, alt text, and captions to provide context.

Example: When including images of "Eiffel Tower" (entity), use filenames like "eiffel-tower-paris.jpg" and provide alt text with relevant descriptive keywords.

12. User-Generated Content:

Encourage user-generated content (UGC) related to entities.

Include features like comments, reviews, and user contributions.

Example: Enable customers to leave reviews for a "hotel" (entity) on your website to generate user-generated content and increase entity relevance.

So what’s this all mean?

Leveraging entity-based ranking techniques, embracing semantic search, implementing entity SEO strategies, and harnessing generative AI, will allow marketers and content creators to stay competitive in the ever evolving search landscape. However, fail to act on this and you’re essetnially leaving it up to the search engines to make sense of who you are and what you’re all about (and trust me, as an SEO, that’s not something you want Google to have free reign on.)

Thank you for reading, be sure to share if this article helped you out! :D