sameAs schema, and why 52% of major sites skip it

By Matt Kerr·Published 6 August 2026

sameAs is the schema.org property that tells a search engine or a language model that this website, that LinkedIn page and that Wikipedia entry are all the same organisation. It is the cheapest entity signal available — a list of URLs you already own — and 52% of the 85 major sites we measured do not have it.

We fetched the homepage of every site in the Docket Index on 2026-08-06, pulled out the JSON-LD, and looked for two things: an Organization-family type, and a sameAs property. 50 sites (59%) declare an organisation. Only 41 (48%) declare sameAs.

The gap between those numbers is the interesting part. 9 sites — 18% of everyone who bothered with Organization schema at all — went to the trouble of describing themselves as an organisation and then omitted the one property that connects that description to anything else.

What the property actually does

A search engine reading your homepage knows a company exists at your domain. It does not know whether that company is the one with 40,000 LinkedIn followers, the one in a Wikipedia article, or the one with a Google Business Profile two miles away. Those are four separate entities until something joins them.

sameAs is that something. It is an array of URLs that are you, elsewhere:

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Acme Roofing",
  "url": "https://acmeroofing.com",
  "sameAs": [
    "https://www.linkedin.com/company/acme-roofing",
    "https://www.facebook.com/acmeroofing",
    "https://en.wikipedia.org/wiki/Acme_Roofing"
  ]
}

Two things follow. Knowledge panels get their social links from resolved entity data, and a language model asked to name a supplier is more likely to reach a business it can pin to a real organisation than one that is only a domain.

Why reference sites are near the bottom

Reference sites came second from bottom at 22%, which looks wrong until you think about who they are. Wikipedia and its peers are the entity graph — they are what everyone else's sameAs points at. They have less to gain by pointing outward.

SEO tools came top at 70%, which is the least surprising number in the table and worth stating anyway: the people who sell advice about this take it.

Getting it wrong

Three failure modes, in rough order of frequency.

Linking profiles in the footer and never declaring them. The commonest by far. The site links Facebook and Instagram from every page, and the schema mentions neither, so the evidence is sitting there unused. Docket reports this specifically because the fix is copying URLs you already have into an array.

Listing pages you do not control. sameAs means "this is also me". A news article about you is not you; that is subjectOf. Padding the array with press mentions weakens the signal rather than strengthening it.

Burying it where nothing looks. Entity markup belongs on the homepage, which is what resolves the site's primary entity. Our measurement only read homepages for exactly this reason — a site declaring sameAs on an About page and nowhere else counts as absent here, and that is also roughly how it looks to something trying to work out who owns the domain.

The other half of the problem

Machines also learn your name from your title tags, your og:site_name and your logo's alt text. When those disagree — and they disagree more often than anyone expects, usually because a tagline crept into one of them — a knowledge panel, a shared link preview and an AI citation can each show a different name for the same company. sameAs connects your entity to the world; consistent naming is what makes the entity coherent in the first place. Docket checks both.

Where another tool is better

For validating that one page's markup parses and is eligible for rich results, use Google's own Rich Results Test. It is authoritative in a way no third-party tool can be, because it is the parser that actually decides. Docket tells you which pages across a whole site are missing the markup and what it costs you; it does not adjudicate eligibility, and anything claiming to is guessing at someone else's parser.

For finding which external profiles and mentions exist to point at in the first place, Ahrefs has a web-scale index and Docket does not. That is a genuine difference in kind, not a feature gap we intend to close.

Checking yours

1. Open your homepage source and search for sameAs. Absent is the common case.
2. List every official profile you actually control.
3. Put them in the array, on the homepage, inside your Organization node — or inside whichever more specific business type you declare, since a subtype is an Organization.

It takes about ten minutes and it is the highest ratio of entity signal to effort available to a small site. The dataset behind this page lists every site measured, so you can check our arithmetic.

How to write the sameAs array

It is one property on the Organization node you already have. If you have no Organization node, that is the first thing to add — sameAs hanging off nothing resolves nothing.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "YOUR ORGANISATION NAME",
  "url": "https://example.com/",
  "sameAs": [
    "https://www.linkedin.com/company/YOUR-COMPANY/",
    "https://github.com/YOUR-ORG",
    "https://en.wikipedia.org/wiki/YOUR_ARTICLE"
  ]
}
</script>

Every value is a placeholder on purpose. A pasteable block with plausible-looking URLs in it is how a factual claim about somebody else's organisation ends up in somebody's markup.

Which URLs belong in it, and which do not

• In: profiles your organisation controls and that represent the organisation itself — LinkedIn company page, GitHub organisation, X account, YouTube channel, Crunchbase, and a Wikipedia or Wikidata entry if one exists.
• Out: anything about you rather than you. A news article, a review, a podcast appearance — those are subjectOf. Padding the array with press mentions dilutes exactly the signal you are trying to send.
• Out: a personal account, unless the node is a Person. A founder's profile does not identify the company.
• Out: profiles you do not control. If you cannot edit it, you cannot keep it true, and a dead or reassigned URL in sameAs is a claim that decays.

Verifying it

Three checks, in order, and the first two are the ones people skip.

1. Is it on the homepage? The homepage is what resolves a site's primary entity, so an array that exists only on an About page is largely wasted. Fetch the homepage itself and look for the block.

2. Is it in the HTML, or only after JavaScript? If the JSON-LD is injected at runtime, the crawlers that do not render never see it — a plain curl of the page shows you what they get. That is the same gap covered in how to run a JavaScript SEO audit.

3. Does every URL still resolve to a profile you control? Open them. This is the part nobody repeats a year later, which is why it is the part that rots.

Docket checks the first two across a whole site and reports the pages where the entity is declared without sameAs, or with an array too weak to resolve anything — part of its AI search visibility lane.

Download Docket

Common questions

What does sameAs do in schema.org?

It lists other URLs that represent the same entity — your LinkedIn page, your Wikipedia article, your Google Business Profile. It lets a search engine or language model confirm that those accounts and your website are one organisation rather than several unrelated ones.

Where should sameAs go?

On the homepage, inside your Organization node. The homepage is what resolves a site's primary entity, so markup on an About page alone is largely wasted.

Should I list news articles about my company in sameAs?

No. sameAs means 'this is also me'. An article about you is not you — that is subjectOf. Padding the array with press mentions weakens the signal.