Entity (SEO)
A uniquely identifiable thing — a person, brand, place, or concept — that a search engine can pin to one distinct record in its knowledge graph.
Last verified · 2026-10-23 · by Moe Ameen
Published 2026-09-03 · Updated
Short answer
In SEO, an entity is a uniquely identifiable thing, such as a person, brand, place, product or concept, that a search engine can pin to one record in its knowledge graph. Google launched its Knowledge Graph on May 16, 2012 under the line "things, not strings." Entity SEO means making your brand easy to recognize and correctly connected to the things around it.
Key takeaways
- An entity is a thing a search engine can resolve to one record, not a keyword string.
- Google launched the Knowledge Graph on May 16, 2012, and Hummingbird followed in September 2013.
- Google says its Knowledge Graph holds over 500 billion facts about 5 billion entities.
- Google says AI Mode draws on the Knowledge Graph, so entity clarity affects AI search too.
- Describe your brand the same way everywhere. Inconsistent descriptions blur the entity.
What is an entity in SEO?
In SEO, an entity is any thing or concept that is singular, unique, well-defined, and distinguishable — a person, a company, a product, a place, an event, an idea, even a color. What makes something an entity is not its category but its resolvability: a search engine can decide, confidently, that a given mention refers to this one specific thing and not to something else that happens to share the name. Entities are the atomic units of meaning in modern search — the recognized things that a keyword string only points at.
The concept exists because search moved from matching strings to understanding things. For most of SEO's history an engine inferred relevance from text patterns and links; it did not really know what a page was about. Google's Knowledge Graph, launched in May 2012 under the slogan "Things, not strings," changed that by holding real-world entities and the relationships between them, so search could reason about a thing rather than just return pages containing a phrase. Entity-based SEO is the practice that follows: optimizing to be understood as a distinct, recognized thing, and to be correctly related to the entities around it, rather than only to match an exact query.
By 2026 this is no longer a fine-grained distinction. The AI systems layered on search lean on entity understanding too (Google says AI Mode draws on the Knowledge Graph), so whether your brand exists as a recognized entity, how it is described, and what it is connected to directly shape whether an answer engine will name you. A page with strong keyword optimization but weak entity signals routinely loses a citation to a page from a brand the engine can actually resolve and trust, because an LLM cannot verify an entity it has no representation for.
When did entities become part of Google Search?
Entity thinking entered mainstream search with two 2012–2013 releases. Google publicly launched the Knowledge Graph on May 16, 2012, describing it as a way to search for things, not strings — a database of real-world entities and their relationships that could power richer, more direct answers. The 2013 Hummingbird update, announced on September 26, 2013 on Google Search's fifteenth anniversary, rebuilt the core ranking algorithm to interpret the intent and context of a whole query rather than weighting isolated keywords, which is what let the Knowledge Graph's entity understanding feed everyday results.
The idea has intellectual roots older than Google's product — semantic-web and linked-data work had modeled the world as entities and relationships for years, and schema.org (2011) gave publishers a shared vocabulary to declare them. What changed over the following decade was scale and stakes: Google has described its Knowledge Graph as holding hundreds of billions of facts about billions of entities, and the arrival of AI answer surfaces trained on that graph turned entity recognition from a knowledge-panel curiosity into the prerequisite for being visible in AI search at all.
How do Google, Bing and AI answer engines handle entities?
| Platform | Behavior |
|---|---|
| Google Knowledge Graph | The canonical entity database behind Search. Confident entities get a stable machine identifier (a MID), and can surface a knowledge panel. Google says AI Mode draws on this graph, so a brand's representation here can shape AI-search visibility. |
| Schema.org / structured data | The vocabulary you use to declare your own entities in JSON-LD. Organization and Person types with a sameAs array explicitly link your entity to its other profiles, handing engines a ready-made corroboration map. Schema does not rank you, but it makes your entity machine-readable and verifiable. |
| Wikidata / Wikipedia | Open, structured knowledge bases that engines lean on heavily to confirm an entity exists and what it is. An entry here is one of the strongest corroboration signals available, but it is earned through genuine notability, not declared — you cannot simply create one for a brand that has done nothing notable. |
| Bing / Copilot | Maintains its own entity graph, known as Satori, separate from Google's. The signals overlap heavily with Google's — consistent naming, corroborating profiles, structured data — even though the underlying graph is separate. |
| LLM answer engines (ChatGPT, Perplexity) | Rank on content quality, specificity, and authority rather than a published entity graph, but still favor sources they can resolve to a recognized, trusted entity. A weak or ambiguous brand entity is a frequent reason a well-written page fails to get named in a generated answer. |
What does entity SEO look like in practice?
- A coach named "James Reed" competes with a UK recruitment firm of the same name. To be recognized as a distinct entity, his site surrounds his name with disambiguating context — his niche, his book, his podcast, his company — so an engine resolves the mention to him rather than the firm.
- A software brand adds a sameAs array to its Organization schema listing its Wikipedia, LinkedIn, Crunchbase, and X profiles. This converts a scatter of separate pages into one confidently-linked record, and the brand starts appearing more consistently in AI Overviews for its category.
- A marketer pastes a draft into Google's Cloud Natural Language API and finds a tangential product name scoring higher salience than the page's actual subject. She rewrites so the intended entity is unmistakably central, tightening the signal about what the page is about.
- A creator standardizes one identical one-line description of their brand across their About page, every social bio, and every guest byline. The consistency is what lets an engine corroborate that all these mentions are the same entity, strengthening the record instead of forcing a judgment call.
What mistakes weaken an SEO entity?
- Treating entity SEO as a replacement for keywords rather than the layer beneath them. You still target queries — you just win them by being a recognized entity with clear topical coverage, not by repeating a phrase.
- Describing the brand slightly differently in every place. Corroboration only works on identical facts; a drifting name, tagline, or category forces engines to doubt whether the mentions are even the same entity, which weakens the whole record.
- Assuming schema markup makes you an entity. Structured data declares and corroborates an entity, but recognition is earned by consistent, verifiable presence across the web — schema that points at thin, inconsistent content buys nothing.
- Trying to fake notability. A Wikipedia or Wikidata entry, third-party coverage, and an established brand SERP accrue from real, notable work; manufacturing them backfires and cannot substitute for substance an engine can verify.
- Building a brand entity around thin content. Recognition earns eligibility to be trusted and cited; it does not manufacture authority. A perfectly consistent entity wrapped around nothing just helps engines confirm, efficiently, that there is little worth quoting.
Is entity SEO a marketing problem or a disambiguation problem?
The most useful reframe of entity SEO is that it is a disambiguation problem before it is a marketing problem. Search does not fail to cite most brands because they lack keywords — it fails because it cannot confidently decide who they are. So the work is unglamorous: pin down one canonical description of the thing and repeat it, identically, everywhere it appears, until an engine has no room left to doubt that all these mentions resolve to one record.
The trap in the AI era is that scaling content with generic tools actively erodes this. If every post drifts in voice and describes the brand a little differently — because a blank prompt returns the model's median house style, not yours — you are feeding engines an inconsistent, blurry entity at high volume, which is worse than publishing less. The point of forcing your specific identity into generation, the way a governed engine like Kompozy does through its Persona Brief and a consistent persona identity, is not aesthetic polish. It is that the brand gets characterized the same way on every surface, so the corroborating signal an engine collects across your output stays coherent instead of dissolving into noise. Consistency is the entity signal; anything that quietly varies your description at scale is working against you.
What does this look like in Kompozy?
A real persona tweet graphic from the Kompozy workspace. Formats like this are generated from one Persona Brief, which keeps a brand described the same way across posts.
Who is this for, and who should skip it?
Best for:
- Brands that share a name with another company, person or product.
- Businesses that want to show up by name in AI Overviews, AI Mode and chatbot answers.
- Teams that publish across many channels and need one consistent brand description.
Not ideal for:
- Thin sites hoping markup alone will earn trust. Recognition comes from real, consistent presence.
- Brands without notable work yet. A Wikipedia or Wikidata entry has to be earned.
- Anyone who wants to drop keyword research. Entities sit under keywords; they do not replace them.
Frequently asked questions
What is an entity in SEO?
An entity is any uniquely identifiable, distinguishable thing — a person, brand, place, product, event, or concept — that a search engine can pin to one unambiguous record in its knowledge graph. Entity-based SEO treats content as being about these recognized things rather than about keyword strings, because Google and the AI systems built on it reason in entities, not exact-match phrases.
How is an entity different from a keyword?
A keyword is a string of characters a searcher types; an entity is the real-world thing that string points at, held with its attributes and relationships. "Apple" is a string that could mean the fruit, the label, or the company — the entity is the specific one an engine resolves from context. Keyword SEO matches the phrase; entity SEO makes you the recognized thing and connects you correctly to related things.
Why do entities matter for AI search?
Google says AI Mode draws on the Knowledge Graph that defines entities, and other answer engines also have to resolve who a brand is, so how your brand is represented as an entity shapes whether an AI answer names you. LLMs cannot verify a brand they have no representation for, so a page with strong keywords but weak entity signals often loses a citation to a page from an entity the engine can resolve and trust.
How do I become a recognized entity?
Give your brand one canonical home page that clearly states what it is; name and describe it identically everywhere it appears; corroborate that identity with consistent profiles and a sameAs array in your Organization schema; and publish substantive content covering your topic and the entities around it. Recognition is earned by corroboration — the same clear description of the same thing in enough trusted places for an engine to resolve it confidently.
Is entity SEO replacing keyword SEO?
No — it sits beneath it. You still target real queries, but you win them by being a recognized entity with clear topical coverage rather than by repeating a phrase. As search shifted from matching strings to understanding things, entity signals became the foundation that keyword relevance now rests on, especially for surfacing in AI-generated answers.
How do we know this?
We checked every figure, date and platform rule on this page against the sources listed below on 2026-10-23, and rewrote or removed anything we could not confirm. Google's own announcements were the primary source for the Knowledge Graph and AI Mode claims.
Sources
- Google — Introducing the Knowledge Graph: things, not strings (read 2026-10-23)
- PBS NewsHour — Hummingbird trumps Caffeine: Google rolls out new search (read 2026-10-23)
- Google Search Central Blog — Introducing schema.org: Search engines come together for a richer web (June 2011) (read 2026-10-23)
- Google — A reintroduction to our Knowledge Graph and knowledge panels (May 20, 2020) (read 2026-10-23)
- Google Cloud — Understand Machine ID (MID) (read 2026-10-23)
- Google — Expanding AI Overviews and introducing AI Mode (Mar 5, 2025) (read 2026-10-23)
- TechCrunch — Bing Now Knows Much More About People And Places Thanks To Its Satori Entity Engine (Mar 21, 2013) (read 2026-10-23)
- Google Cloud — Natural Language API Basics (read 2026-10-23)
Related terms
- Generative Engine Optimization (GEO) — The practice of shaping content so AI answer engines like ChatGPT, Perplexity, and Google’s AI Overviews cite and quote it in their generated answers.
- Schema markup for AI citations — Using schema.org structured data to make a page machine-readable and its entity verifiable for AI answer engines — an assist to citation, not a cause of it.
- Content anchoring — Content anchoring is organizing every post around one central asset, theme, or buyer belief, so each piece reinforces the same idea instead of drifting.
- Brand messaging — The strategic set of core messages — value proposition, positioning, message pillars, and tone — a brand repeats consistently across every channel.
- AI Content Repurposing — The complete methodology for turning one source into 25-35 pieces of native-format content across every platform — without producing AI slop.
- Autonomous Content Creation — Most "autonomous" AI content is slop.
- AI Brand Voice & Persona — Without a Persona Brief, every AI output averages to the LLM default voice.