How to Find Entities for SEO Optimization: Forget Keywords First

An entity, in the context of SEO and Google's Knowledge Graph, is any distinct, real-world concept that can be uniquely identified and differentiated from other concepts. People, places, organisations, products, events, abstract ideas, and creative works can all be entities. What makes something an entity is not the word used to describe it — it is the fact that Google has given it a stable identity that persists across different phrasings, languages, and contexts.

Google describes entities in its own documentation as 'things, not strings.' That phrase contains the entire philosophy shift you need to make. A string is 'best running shoes.' An entity is the concept of 'running footwear' as understood in relation to athletics, biomechanics, specific brands, and consumer intent. Google does not match your content to a query by checking for string overlap.

It checks for conceptual alignment between the entity space of your content and the entity space of the query.

This is why two pages using completely different vocabulary can rank for the same query — and why two pages using identical keyword density can rank for entirely different things. Entity identity is determined by the full ecosystem of concepts present in and around your content, not by keyword frequency.

For practical SEO purposes, entities fall into several broad categories:

- Named entities: specific people, places, organisations, products, events - Conceptual entities: ideas, disciplines, methodologies (e.g., 'machine learning,' 'content marketing') - Attribute entities: characteristics that define other entities (e.g., 'open-source,' 'peer-reviewed') - Relational entities: concepts defined primarily by their relationship to other entities (e.g., 'subsidiary,' 'inventor of')

When you find entities for SEO optimisation, you are not looking for words to add. You are mapping the conceptual landscape Google expects to see when it reads content claiming authority on a topic. That map determines your content architecture, your internal links, your schema choices, and your author positioning — not just your on-page copy.

The Entity Gap Audit is the framework we use at the start of every content strategy engagement, and it consistently surfaces opportunities that keyword gap analysis completely misses. The premise is simple: if a competitor is consistently ranking above you for a cluster of related queries, it is often because their content has established entity relationships your content has not — even if your keyword coverage appears comparable.

Here is the Entity Gap Audit process in full:

Step 1: Identify your target entity and its top-ranking competitors Choose the central entity your page or site section is meant to represent. Find the top 3-5 ranking pages for your primary target query. These are your entity benchmarks.

Step 2: Run each competitor URL through Google's Natural Language API Google's NL API (available free at the Cloud Console demo) returns a salience-ranked list of entities it detects in any piece of content, along with entity type, Wikipedia URL where applicable, and a salience score indicating how central each entity is to the document. Run your own URL and each competitor URL through this tool.

Step 3: Map the entity overlap and gaps Create a simple matrix: your page's entities in one column, each competitor's entities across additional columns. Entities that appear consistently in competitor content but are absent from yours are your entity gaps. High-salience entities that you are missing entirely are your highest-priority additions.

Step 4: Distinguish between entities and noise Not every entity the NL API surfaces is intentional or valuable. Look for entities that appear with both high salience and Wikipedia/Wikidata links — these are confirmed Knowledge Graph entities, not just named noun phrases the API is guessing at.

Step 5: Prioritise by relationship type Some missing entities are sub-topics (entities that sit below your main entity in the knowledge hierarchy). Others are co-occurring context entities (entities that signal which category your main entity belongs to). Prioritise co-occurring context entities first — they establish your entity's identity.

Sub-topic entities expand depth once identity is established.

What makes the Entity Gap Audit more powerful than keyword gap analysis is that it reveals structural differences in how content is understood, not just surface-level vocabulary differences. You might discover that a competitor's page is winning not because of better keywords, but because it mentions three foundational context entities that anchor the page's identity in exactly the topic category Google associates with that query intent.

One of the hardest parts of entity-based content planning is knowing where to look. The PEAT Stack framework — People, Events, Attributes, Things — is a categorisation system we developed to make entity discovery systematic rather than ad hoc. It ensures you discover entities across all relevant dimensions of your topic, not just the most obvious ones.

P — People For almost every topic, there are named individuals whose association with the subject matter signals authority and topical depth. These might be researchers, practitioners, founders, historical figures, or thought leaders. When Google sees content about a discipline that correctly identifies and contextualises the key people associated with that discipline, it is receiving strong confirmation of entity identity.

Ask: who are the established authorities, founders, or key practitioners in this topic's Knowledge Graph? Include them by name with correct context.

E — Events Events are often underused entity signals. Conferences, product launches, research publications, historical milestones — these are entities with specific Knowledge Graph identities that, when correctly referenced, place your content in a verified temporal and contextual framework. For example, content about machine learning that correctly references the ImageNet competition, the Turing Test, and AlphaGo's 2016 victory is signalling entity-level depth that keyword-matched content cannot replicate.

A — Attributes Attributes are the properties and characteristics that define your main entity. These are often where content falls shortest. What are the defining characteristics, sub-types, measurement dimensions, or qualitative properties associated with your primary entity?

Attributes anchor your entity in its correct semantic neighbourhood. For a page about 'content marketing,' attributes might include 'long-form content,' 'editorial calendar,' 'audience segmentation,' and 'content distribution.' These are not keywords — they are definitional properties of the entity.

T — Things This is the broadest category: tools, products, platforms, methodologies, frameworks, and physical objects that are part of your entity's ecosystem. These are often the easiest entities to identify but the most commonly over-used. Mentions of Things without People, Events, or Attributes produce a content profile that Google may associate with commercial intent rather than topical authority.

The PEAT Stack works best as a discovery checklist before writing, not as an editing pass after. Build your PEAT inventory for a topic, then let it shape your content outline. A page with strong representation across all four categories reads as authoritative not just to humans, but to a Knowledge Graph that expects to see this full dimensional picture of a legitimate topic.

Keyword tools are built to surface search demand, not entity relationships. Using a keyword tool to find entities is like using a road atlas to understand topography — you get surface-level geography without the structural depth. The best entity discovery happens in sources that reflect how Google's Knowledge Graph is actually built.

Wikipedia and the Topic Graph Wikipedia is one of Google's primary entity sources. Every link within a Wikipedia article represents an explicit entity relationship. Open the Wikipedia article for your primary topic entity and study three things: the categories at the bottom of the page (these reveal how Google classifies your topic), the infobox attributes on the right side (confirmed entity properties), and the 'See also' section (related entities at the same hierarchical level).

This is not generic research — it is reading the entity graph directly.

Wikidata as a Structured Entity Map Wikidata is Wikipedia's structured data layer, and it is the closest thing to a public map of Google's Knowledge Graph. Every entity in Wikidata has a unique identifier (Q-number), a set of properties, and explicit relationships to other entities. Search your primary topic entity on Wikidata and examine its full property list.

Properties with 'related entities' values are your PEAT Stack inputs, ready-catalogued.

Google Knowledge Panels Search your primary entity on Google and study the Knowledge Panel that appears. The 'People also search for' section, the categorisation label (e.g., 'Computer scientist,' 'Marketing strategy'), and the attribute rows displayed are all direct signals of how Google understands this entity and its relationships. This is the entity map Google will use to evaluate your content.

Google's Natural Language API Beyond the competitor audit use case described earlier, run the NL API on the top-ranked Wikipedia article for your topic. Wikipedia articles are entity-rich by design, and the NL API output on a Wikipedia page gives you a curated, salience-ranked entity inventory that Google itself uses as a reference point for the topic.

Search Console's Query Clusters Search Console does not directly surface entities, but query clusters reveal entity intent. Group your existing queries by conceptual theme, not keyword similarity. Queries clustering around the same conceptual territory indicate an entity that your page is partially satisfying — a signal to develop that entity more fully.

Google's 'People Also Ask' and Related Searches These features are entity relationship signals in disguise. PAA questions reveal the attributes and sub-entities users associate with your primary entity. Related searches reveal co-occurring entities.

Mine both systematically for any primary entity you are targeting.

The single biggest opportunity most SEOs leave on the table in entity optimisation is off-page entity signals. Google does not determine your domain's entity authority from your content alone. It builds a picture from every entity signal across the web — your brand mentions, your author profiles, the entity context of your backlinks, and the entity language used in anchor text and surrounding content.

Brand as Entity Your brand name is an entity. If Google has a Knowledge Panel for your brand, that is confirmation of entity status. If it does not, your entity authority is operating with a significant handicap — Google is less certain of who you are, which makes it less certain about the authority you claim.

Building brand entity status means creating and maintaining consistent entity signals across multiple platforms: your own site (with proper Organisation schema), Google Business Profile, LinkedIn, relevant industry directories, and earned media mentions where your brand name appears in contextually relevant entity neighbourhoods.

Author as Entity Every author on your site is an entity claim. When a person publishes content under their name, Google evaluates whether that person has an established entity identity consistent with the expertise claimed. This is the EEAT dimension of entity SEO.

Author entities need Wikipedia or Wikidata presence in high-authority spaces, or failing that, consistent entity signals across professional platforms, bylines in established publications, and well-structured author schema on your own site. We have seen pages with identical content perform differently based solely on whether the author's entity is established or anonymous.

Backlink Context as Entity Signal Not all backlinks contribute equally to entity authority. A backlink from a page whose entity space closely overlaps with yours transfers not just PageRank but entity authority. A link from a page about 'content strategy' to your page about 'editorial planning' in a content marketing context carries entity relevance that a generic directory link cannot replicate.

When prospecting for links, prioritise pages whose entity profile (run them through the NL API) overlaps with your target entity space.

Anchor Text as Entity Confirmation Anchor text in backlinks is one of the most direct entity signals Google receives from third parties. Exact-match anchors for your target entity are powerful precisely because they confirm external parties' understanding of what your page is about. Diversified anchor text that uses entity synonyms and related entities is more natural and often more valuable than keyword-precise anchor text that looks engineered.

Structured data is where entity optimisation most frequently gets misapplied. The common advice — 'add schema markup to signal entities to Google' — implies that schema is how you establish entity identity. In practice, schema is how you confirm entity identity that already exists in your content.

Getting this sequence wrong wastes implementation effort and sometimes creates contradictory signals.

Think of it this way: if your prose content clearly establishes that a page is about a specific person — naming them, describing their attributes, contextualising their relationships to other entities — then Person schema confirms and disambiguates what Google has already understood. If your prose content is entity-ambiguous but you add Person schema, Google is receiving a claim without evidence. In low-competition spaces, this might still work.

In contested topic areas, content without prose-level entity establishment rarely wins on schema alone.

The Entity Confirmation Sequence 1. Establish your primary entity clearly in the first 100 words of content — name it, describe a defining attribute, and contextualise it in relation to one other confirmed entity 2. Develop entity depth through the body of the content using your PEAT Stack inventory 3.

Apply structured data that matches the entity type you have established in prose 4. Use sameAs properties in schema to link your entity to its Wikipedia, Wikidata, or other authoritative identifiers — this is the most underused and most valuable schema technique for entity confirmation

The sameAs Property The sameAs property in Schema.org markup is the most direct entity signal you can give Google through structured data. By including the Wikidata URL and Wikipedia URL for your primary entity in an Organisation, Person, or other schema type, you are explicitly linking your content's entity claim to a Knowledge Graph node Google already trusts. This is entity confirmation in its most precise form.

Schema Types That Directly Support Entity SEO - Organisation with sameAs: for brand entity establishment - Person with sameAs: for author entity establishment - Article with author referencing a confirmed Person entity: for content entity authority - FAQPage: for entity attribute coverage through question-answer format - HowTo: for process entity establishment - BreadcrumbList: for entity hierarchy signalling through URL structure

The most powerful schema implementations link multiple entity types together. An Article whose author is a confirmed Person entity, published by a confirmed Organisation entity, about a topic with sameAs links to Wikidata — this creates an entity graph in your schema that mirrors the Knowledge Graph relationships Google already understands.

Internal linking is taught almost exclusively as a crawlability and PageRank distribution tool. Both of those functions are real. But internal linking also serves a third function that rarely gets discussed: entity reinforcement.

The entity signals present in and around your internal links — anchor text, surrounding sentence context, the entity profile of the linking page, and the entity hierarchy implied by your URL structure — collectively tell Google how your site's entity space is organised.

When we restructure internal linking architecture with entity relationships as the primary organising principle, the results in topical authority building are consistently more durable than when linking is organised purely by keyword targeting or PageRank flow. Here is why: Google builds entity associations at the domain level partly by observing which entities consistently appear in proximity and link to each other across your site. A site where entity-relevant pages consistently link to each other — with entity-accurate anchor text and in contextually coherent surrounding copy — builds an entity graph that mirrors Google's own Knowledge Graph for that topic.

Entity-Led Internal Linking Principles

Principle 1: Anchor text should use entity names, not keyword phrases When linking to your page about 'editorial calendars,' the anchor should be 'editorial calendar' (the entity name) rather than 'best editorial calendar practices for content teams' (a keyword phrase). Entity names as anchors are more precise signals of what the linked page is about.

Principle 2: The surrounding sentence context matters Google reads the sentence surrounding an internal link as context for understanding why these two pages are connected. A link to your editorial calendar page that appears in a sentence about 'structuring your content production workflow' carries entity relationship context that a link in a generic 'see also' list does not.

Principle 3: Map your internal links against your entity hierarchy Your pillar pages should represent primary entities in your topic space. Supporting pages should represent sub-entities or attribute entities. The internal linking between them should flow in a way that reflects the entity hierarchy in your PEAT Stack — pillar entities linking to and from attribute entities, event entities, and people entities that populate the broader topic graph.

Principle 4: Avoid orphaned entity pages A page about a topic entity that receives no internal links is an entity without confirmation from its own domain. It exists in isolation, unable to benefit from or contribute to the domain's entity graph. Every entity page you create should be linked from at least two contextually relevant pages on your site.

Entity optimisation is not a publish-and-move-on exercise. The entity landscape for any topic evolves — new entities enter the Knowledge Graph, existing entities gain or lose prominence, and your competitors' entity coverage changes. A monitoring and iteration cadence is what separates entity SEO as a short-term tactic from entity SEO as a compounding authority-building system.

What to Monitor

The first monitoring layer is your Search Console query data, analysed through an entity lens. Rather than tracking individual keyword rankings, group your queries by the entity they represent. A query cluster that is growing in impressions but not in clicks suggests an entity Google associates you with but has not yet confirmed you are authoritative about.

A cluster growing in both impressions and clicks confirms entity authority building is working.

The second layer is Knowledge Panel tracking. Search your brand entity, your primary topic entities, and your key author entities quarterly. Changes in Knowledge Panel attributes, related entities, or category labels tell you how Google's understanding of your entity space is evolving — sometimes reflecting your optimisation work, sometimes revealing gaps you need to address.

The third layer is competitor entity audits. Re-run your Entity Gap Audit against evolving competitor content every quarter. New entities appearing in competitor pages at high salience, or shifts in which entities are most salient across the competitive set, are early indicators of where Google's entity weighting for the topic is moving.

The Compounding Mechanism Entity SEO compounds because entity authority is self-reinforcing. As your domain establishes entity authority in a topic space, new content you publish in that space inherits some of that authority — it is indexed faster, ranked higher at launch, and cited more readily by Google's AI-powered features. The entity infrastructure you build on your first ten pieces of content makes your next fifty pieces more efficient to produce and more effective to rank.

This is why entity optimisation is fundamentally a long-game strategy. The sites that dominate topic spaces in competitive niches are not those with the most content or the most backlinks — they are those with the most coherent, comprehensive, and confirmed entity graphs across their entire domain. Building that graph systematically, auditing it regularly, and expanding it deliberately is the highest-leverage SEO activity available to any site operating in an information-dense market.