SEO Pages

Entity SEO: how Google learns your business is a real thing

Written by

Personally reviewed

Founder, SEO Pages

Entity SEO teaches Google and AI tools that your business is one identifiable thing. The nine checks we run, with the findings from our own site.

Published

Add SEO Pages as a preferred source on Google
Table of contents
A storefront building with a glowing outline connected by lines to verification symbols

What is an entity in SEO?

An entity is a thing with an identity: a specific person, place, company, product or idea that exists apart from the words used to describe it. A string is only the text. "Hartley Plumbing" typed into a search box is a string. The plumbing firm in Bromley with a phone number and a Companies House record is an entity.

Strings are ambiguous. Two firms can share a name, and one firm can go by three. A search engine that only matches strings cannot say which Hartley you meant. One that models entities knows there is one firm, with several names, doing plumbing in Bromley.

Entity based SEO is building your site and web presence so identity is obvious: you state who you are once, in a form machines read, and every page and profile agrees.

Entity SEO: moving from strings to things

Google made the shift official in 2012, when it launched the Knowledge Graph under the line "things, not strings". Amit Singhal, then Google's senior vice president of engineering, explained the idea with a query: [taj mahal]. For more than four decades, he wrote, search "has essentially been about matching keywords to queries", so to a search engine [taj mahal] was two words.

The words could mean "one of the world's most beautiful monuments, or a Grammy Award-winning musician, or possibly even a casino in Atlantic City". Google described building "an intelligent model" that "understands real-world entities and their relationships to one another". Once it did, it could ask which one you meant: "Do you mean Taj Mahal the monument, or Taj Mahal the musician? Now Google understands the difference."

That is the whole principle. The same string points to different things, and the system needs a way to tell them apart. A local business has the same problem on a smaller scale. A "Hartley Plumbing" in Bromley and another in Leeds share a string. They are different entities.

What is Google's Knowledge Graph today?

Google's Knowledge Graph is, in Google's own words, "our database of billions of facts about people, places, and things." Google's Knowledge Panel help page says it uses the graph to answer factual questions and aims to "discover and surface publicly known, factual information when it's determined to be useful."

In 2012 Google reported more than 500 million objects and more than 3.5 billion facts about them. The current help page says "billions of facts".

The same page says where facts come from: "a variety of sources that compile factual information", licensed data for things like sports scores and weather, and information "directly from content owners in various ways". It also says that knowledge panels "are created automatically by Google Search Algorithm when there is enough information available on the open web." You do not apply for a place in the graph. You can make the record easier or harder to build.

Google's knowledge panel overview says businesses "have special Business Profile panels and should use Business Profile to create, claim, and make suggestions to those." Our guide on how to get a Google knowledge panel covers which panel is which, and the knowledge panel checker shows whether Google recognises your business or name as an entity.

Why entities matter for a local business and for AI answers

Entities matter because a search system can only show, rank or recommend a business it has identified. If your name, address and services appear in slightly different forms across your site and your profiles, a system has to guess whether they describe one business or several.

There is a second reason. When someone asks an AI tool "Who is the best emergency plumber in Bromley?", the tool has to name particular businesses and say something true about each. This is our reasoning, not a statement from Google: a tool can only recommend a business it can identify, so a business with a clear identity, consistent facts and corroborating sources is easier to name than one with scattered, conflicting details. We cover how this plays out in what is AEO.

Two things Google's own documents say keep this honest. Google's AI features guide says: "There's also no special schema.org structured data that you need to add" to appear in AI Overviews or AI Mode. And its structured data policies say Google "does not guarantee that your structured data will show up in search results". Entity work makes the facts clear and true. It guarantees nothing.

Google does say that organisation markup on your home page "can help Google better understand your organization's administrative details and disambiguate your organization in search results" (Organization structured data). Disambiguate is the entity word for "tell apart from others".

The nine checks we run

We run nine checks on a site's entity layer. We ran them on seopages.co.uk on 20 October 2026, fixed every finding the same day and re-checked all 199 pages in the sitemap afterwards. An earlier audit on 25 August 2026 had found a thin entity home, provider stubs that created a fresh anonymous organisation on each page, and served places with no anchor to say which town was meant.

CheckThe question it answers
A. Entity identityIs there one canonical node for the business?
B. Graph connectivityDoes one thing have one @id?
C. Knowledge-base anchoringDo sameAs links point to pages that confirm who you are?
D. AliasesAre alternative names declared in markup?
E. Attribute coverageDoes the markup answer what people ask about the thing?
F. Subject entitiesDoes each page name what it is about?
G. ConsistencyIs there one name everywhere?
H. Topical focusDoes the site stay on the subjects it claims?
I. Honesty gatesIs every entity claim true?

JSON-LD is a block of structured data in a page's code that describes the page in a format machines read. Schema.org is the shared vocabulary of types and properties it uses. An @id is a unique web address-style label that names one thing inside that data. Our guide to schema markup explains the basics.

A. Entity identity: one canonical node

Check A asks whether your site declares one main node for the business, usually on the home page, with the facts that make it identifiable. Look for a legal name, a registration number, an address, a logo and a stable @id.

Here is the start of our own Organization node, trimmed from the live JSON-LD on our home page:

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://www.seopages.co.uk/#organization",
  "name": "SEO Pages",
  "legalName": "SEO Pages Ltd",
  "identifier": {
    "@type": "PropertyValue",
    "propertyID": "GB-COH",
    "value": "14129018"
  }
}

The @id gives the business a permanent label. The legalName separates the company from the trading name. The Companies House number, declared as an identifier with the property ID GB-COH, is a fact a machine can check against a public register, and no other company shares it.

The August 2026 audit found our home page was a thin entity home: it carried too few of these facts. The live node now carries all of them.

B. Graph connectivity: one thing, one @id

Check B asks whether every mention of the same thing uses the same @id. A graph is only a graph if the nodes connect. If every page invents its own label for your business, a machine sees a crowd of strangers instead of one company.

Two of our findings sat here. In the August audit, provider stubs on service pages created a fresh anonymous organisation on each page. Each page said "this service is provided by an organisation" without pointing to the real one. In the October audit, one @id carried two different names for the same service. That is a contradiction inside our own data: one label, two descriptions.

The fix is to define each thing once and refer to it by @id everywhere else. Our service nodes now name the provider like this:

"provider": {
  "@type": "Organization",
  "@id": "https://www.seopages.co.uk/#organization",
  "name": "SEO Pages"
}

Every service we sell has its own @id and points back to the one Organization node.

C. Knowledge-base anchoring with sameAs

Check C asks whether your markup links to other places on the web that confirm who you are. The property for this is sameAs. Google's Organization guidance describes it as "the URL of a page on another website with additional information about your organization", such as a profile on a social media or review site, and says you can give several.

Anchoring ties your entity to records that other systems already hold. Here is a trimmed version of our list: four ordinary profiles, then the Google Knowledge Graph link:

"sameAs": [
  "https://www.linkedin.com/company/seo-pages",
  "https://www.facebook.com/SEOPages/",
  "https://www.youtube.com/@seopagesofficial",
  "https://clutch.co/profile/seo-pages",
  "https://www.google.com/search?kgmid=/g/11zggns48z"
]

The kgmid link points to our entry in Google's own graph. Our live list is longer, and each entry has to be a page that exists and describes us.

This check produced our most useful finding. Our blog posts declare Google Business Profile as a subject, and its anchor pointed at a Wikipedia title that redirects into the Google Maps article. A redirect like that can merge two different things: our markup meant Google Business Profile, the product, and the link resolved to Google Maps. We changed the anchor to the product's own Wikidata item (Q21235252) instead.

The lesson applies to any site. Open every sameAs link in a browser and check where it lands. If a Wikipedia address redirects, the entity on the other end may not be the one you meant. If a profile has been deleted, the link vouches for nothing. A separate post on Organization schema and sameAs is coming, with the full implementation.

The August audit also found a problem in the same area: served places with no anchor to say which town was meant. A page for "Bromley" is a string until something ties it to the Bromley you serve.

D. Aliases declared in markup, not stuffed in copy

Check D asks whether the other names people use for you or your services are declared as data. Schema.org has alternateName for this. Our Organization node carries "SEOPages" as an alternate name, and each service lists the phrases people use for it.

Keep aliases out of your page copy. A plumber who is also "Hartley's Heating" can say so once in markup and leave the visible text natural.

E. Attribute coverage: answering what people ask about the thing

Check E asks whether your markup holds the facts people want to know about the entity. For a business that usually means what it does, where it works, how to contact it, when it started and how it differs from similarly named things.

Two properties from our own node show the idea. knowsAbout lists the subjects the business covers. disambiguatingDescription says in one sentence what the entity is and what it is not. Ours reads:

"disambiguatingDescription": "SEO Pages is a UK digital marketing agency (SEO Pages Ltd, Companies House 14129018), not the generic term for search-optimised web pages."

We needed that sentence because our name is also a phrase. "SEO pages" describes a kind of web page, so a machine reading our name alone has two readings. One sentence resolves it. If your name collides with a common word, a place or another firm, write the equivalent for yours.

F. Subject entities on content

Check F asks whether each page tells machines what it is about. Schema.org has about for the main subject and mentions for things the page names in passing. The rule is to list only what the page discusses. A page about boiler servicing can say it is about "boiler servicing" and mentions "Gas Safe Register". It cannot list ten loosely related topics to look broader.

G. Consistency: one name everywhere

Check G asks whether the same name, labels and facts appear on every page, in the markup and in the visible text. Inconsistency is the most common entity fault on small sites, and it comes from templates.

Ours was a shared fallback. When a local area page had no service label of its own, the template filled in a default, and that default labelled 25 local area pages as "Programmatic SEO", a service we do not sell under that name. Nobody wrote those labels by hand. One template fallback produced all 25.

The same check covers the outside world: name, address and phone number should match across your site, Google Business Profile and listings. Our guide to local business listings and consistent details covers that half.

H. Topical focus

Check H asks whether the site stays on the subjects the entity claims to be about. If your markup says you are a plumber, your pages should be about plumbing and the places you serve. An entity that claims ten unrelated specialities is harder to describe than one that claims two.

I. Honesty gates: entity claims must be true

Check I asks whether every claim in your markup is accurate. Entity markup states facts about your business in a form machines read. A claim that is false, outdated or inflated makes the whole record less reliable.

Our own finding: the free audit page described itself as covering paid ads when it does not. The markup repeated the claim. We corrected both.

If you cannot show a claim to a customer, do not tell a machine.

What a fixed entity layer looks like

After the fixes, one Organization node holds the identity. Services each have one @id and point back to it. The founder is a Person with a separate @id. Profile links and the Knowledge Graph link anchor the whole thing to records outside our control.

Two caveats. We cannot show that any of this moves a ranking, and Google does not say it does. And the nine checks are our method, not a Google standard: Google documents the building blocks (structured data, sameAs, the Knowledge Graph) and the grouping is ours.

You can check part of your own site today. Paste a page into our schema validator to see what JSON-LD it carries and whether any of it is invalid. Run the knowledge panel checker to see whether Google recognises your business or name as an entity. Then ask the AI tools directly: the AI visibility checker puts questions to them and shows which businesses they name.

Where to start on a small site

A small site does not need nine checks run in order. These five steps cover most of the gain, in the order that avoids rework:

  1. Fix your facts first. Settle one business name, address and phone number, and make your site, Google Business Profile and listings match (check G).
  2. Add one Organization or LocalBusiness node to the home page. Give it a stable @id, your legal name and, for a UK company, your Companies House number (checks A and E). The schema generator writes the JSON-LD.
  3. Add sameAs links, and open every one. Include only profiles that exist and describe your business. Check where each link lands (check C).
  4. Point every other page back to the one node. Service pages should refer to your @id as the provider instead of inventing a new organisation (check B).
  5. Re-check after every template change. Our 25 mislabelled area pages came from one shared fallback. Run the validator again whenever a template changes (checks G and I).

Leave about and mentions until the first four steps are done.

Entity SEO sits inside the wider technical groundwork in our technical SEO guide, alongside crawling, indexing and speed. Entity work sits on top of those basics.

FAQ

What is an entity in SEO?

An entity is a distinct thing, such as a business, person or place, that a search system can identify and tell apart from other things with similar names. It differs from a keyword, which is just text. A business becomes an entity in practice when its name, facts and profiles agree with each other across the web.

What is the difference between entity SEO and keyword SEO?

Keyword SEO matches the words on a page to the words in a search. Entity SEO makes sure the system knows which real thing those words refer to. Keywords decide what you are found for; a clear entity decides whether you are identified correctly.

Does schema markup make Google recognise my business as an entity?

Schema markup helps Google understand your pages and can help it tell your organisation apart from others, according to Google's Organization guidance. It does not guarantee a knowledge panel, a rich result or an AI mention. Google builds Knowledge Graph records automatically "when there is enough information available on the open web".

Do I need a Wikipedia page for entity SEO?

No. Wikipedia is one place a business can be recorded, and most small businesses do not meet its notability rules. Use sameAs links to profiles that exist and describe you, such as LinkedIn, your Google Business Profile, trade directories and Companies House. Check each link lands on the right page, because a redirect can point at a different entity.

How do I check whether Google knows my business as an entity?

Search for your business name and see whether a knowledge panel appears. Then run our Google knowledge panel checker, which tests whether Google recognises a business or name as an entity and gives a readiness score. Google decides when a panel shows.

If you want a business that AI tools can identify and name, we build and fix this layer for clients. See how our AI visibility service works.

SEO Pages assistant

Ask us anything. A human reads every chat.

Hi, I'm the SEO Pages assistant. Ask me anything about getting your business found in Google and AI search.
Book a free call →