Entity SEO: making search engines understand your brand

Entity SEO is the work of making search engines and AI assistants recognize your brand as one specific thing, with one name, one website and a set of profiles that all agree. You do it by using the same name everywhere, stating it in Organization markup with an @id and sameAs links, keeping your public profiles consistent and, if you qualify, having a Wikidata item that points to your site.

What is an entity in SEO?

An entity is a distinct, identifiable thing such as a company, a person, a product or a place, as opposed to the string of letters that names it. Google introduced its Knowledge Graph in a May 2012 post with the line "things, not strings", and said it held over 500 million objects and 3.5 billion facts about them at launch.

The difference matters most when your name is ambiguous. "Juniper" is a shrub, a networking company, a bank and a dozen startups. A page that says "Juniper" tells Google nothing on its own. A page whose markup, title, profiles and outside mentions all say "Juniper Ledger, bookkeeping software for dental practices, juniperledger.com" gives Google one thing to file it under.

Why AI answers need a clear entity

ChatGPT, Gemini and Perplexity describe a brand by combining what they find about it. If your signals disagree, the model has to guess which facts belong to you, and it may merge you with a namesake or leave you out of a recommendation because it isn't sure you exist as one company.

Renamed brands hit this most often. The old name lives on in a Crunchbase profile and half the review sites, so an assistant can describe the product under its old name with old facts. Nothing on the current site is wrong. The entity has split in two.

A clear entity also makes your own content count. Answers that cite you build on a description the model already holds, and getting mentioned by AI assistants is far easier when every mention points to the same thing. If you want to see what the engines currently say, Brand Narrative Check asks ChatGPT, Perplexity and Gemini directly.

Use one brand name everywhere

Pick one public name and repeat it exactly in every place a machine reads it. Those places are your Organization name, WebSite markup, og:site_name, the home page <title>, the web app manifest, llms.txt if you have one, and the display name on every profile.

Spelling counts. "Juniper Ledger", "JuniperLedger" and "Juniper Ledger Inc." look the same to a person and different to a parser. Put the registered company name in legalName and a real short form in alternateName, so variants are declared instead of scattered.

Then fix the profiles, not just the site. The usual offenders are an old LinkedIn page, a Crunchbase entry someone filled in at seed stage, a G2 listing with the previous logo, and a GitHub organization with no website link. Each one should carry the same name, logo, one-line description and a link back to your home page. Links in both directions are what let a crawler confirm the match.

Mark up your brand entity with Organization schema

Organization markup is where you state the entity in a form engines read directly. Put it once on your home or About page, give it a stable @id, and reference that @id from everywhere else instead of repeating the block.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://www.juniperledger.com/#organization",
  "name": "Juniper Ledger",
  "url": "https://www.juniperledger.com/",
  "sameAs": ["https://www.linkedin.com/company/juniper-ledger"]
}

An Article's publisher or a Person's worksFor then points to {"@id": "https://www.juniperledger.com/#organization"}, and every page on the site names the same node. Which properties to add and the logo rules are in the Organization schema guide. Which profiles belong in sameAs, and which template placeholders to delete, are in the sameAs schema post. For the people behind the brand, author schema applies the same idea to Person markup.

Wikidata and checking what Google already knows

Wikidata is a public knowledge base where each entity gets an ID, and an item whose official website property (P856) lists your home page is a strong, open link between your name and your domain. Add the item's URL to your sameAs so the link runs both ways.

You can't just create one for a brand with no outside coverage. Wikidata's notability rules expect serious independent references, and the knowledge panel guide covers who qualifies and how to handle Wikipedia without breaking its conflict of interest rules.

To see whether Google already holds an entity for you, search your exact brand name and look for a panel. Developers can query the Knowledge Graph Search API, which returns matching entities by name. Google says it isn't meant for production-critical use, so treat it as a spot check, not a monitor.

Check your brand entity signals

Brand Entity Checker reads your home page and compares the brand name across your Organization and WebSite markup, og:site_name, the application-name tag, the manifest, llms.txt and the page title. It checks that the Organization markup has a name, a url on your site, a description and a logo that loads at a usable size. It opens up to ten sameAs links and flags dead profiles, network home pages left by templates and links back to your own site. It also looks for a Wikidata item that lists your site, and tells you when sameAs doesn't link that item or its Wikipedia article.

Each finding comes with the fix. A run of the brand entity checker costs 8 credits. It doesn't create Wikidata entries, monitor mentions on other sites or tell you what AI answers say about you.