sameAs Schema: How a Site Proves It Is Who It Claims

sameAs is the schema.org property that lists other pages describing the same entity as yours: a Wikidata entry, a Wikipedia page, an official profile, a parent company. Its only job is entity disambiguation, helping a machine tell your organization apart from every other one carrying a similar name. It's a claim you make about yourself, not a verification you receive.

What Does sameAs Actually Declare?

sameAs declares that a target URL "unambiguously indicates the item's identity, e.g. the URL of the item's Wikipedia page, Wikidata entry, or official website" (schema.org, 2026). That word, unambiguously, is the whole point: the property exists to remove doubt about who an entity is, not what it does. It's defined on schema.org's Thing type, not Organization specifically, so it applies to any entity. The expected value is a URL, nothing more.

Google's own description matches: sameAs is "the URL of a page on another website with additional information about your organization," and a site "can provide multiple sameAs URLs" (Google Search Central, 2026), marked Recommended, not required. It's not a link that passes authority or a recommendation, just a machine-readable identity assertion, worth understanding before touching how structured data fits into AI visibility.

Why Does Identity Ambiguity Break Things?

A machine reading your site has no built-in way to know that a name and an entity are two different things. Two organizations can share an identical name with zero connection between them, and a single organization can operate under several names across markets. Nothing forces those to resolve on their own.

Google confirms this is a real design concern: some Organization properties are "used behind the scenes to disambiguate your organization from other organizations" (Google Search Central, 2026). When that disambiguation fails, attribution bleeds across entities, an activity or a fact getting credited to the wrong organization because nothing separated the two clearly enough. That's precisely when Google mixes your entity up with another one, a case worth reading on its own.

How Do You Choose Which URLs Belong in sameAs?

Choosing what goes into sameAs isn't about picking URLs that flatter you, it's about picking URLs that let a machine resolve who you are. None of the pages currently ranking on this topic give a reproducible way to make that call, so here is one.

Does the page identify the entity?

A candidate URL has to name and describe your organization specifically, not merely exist somewhere with your name in it. A directory listing that mentions you in passing doesn't identify you; a Wikidata item built around your organization does.

Does it point back?

sameAs is a one-way claim about a page you don't control. Nothing corroborates it unless that page also links back to you, the part almost nobody on this topic states plainly.

Is the target itself an identified entity?

A Wikidata entry or an official business registry carries more weight than an abandoned social profile, because the target's own identity has already been resolved by someone else.

That gives a simple hierarchy: authority databases first (Wikidata, Wikipedia, official registries), controlled profiles that link back second, everything else doesn't belong. Wikidata isn't a form you fill out: an item is accepted only if it meets specific notability criteria (Wikidata, 2026). schema.org sets no numeric limit, and Google confirms multiple URLs are fine (Google Search Central, 2026), so the constraint was never quantity, it's the trust signals Google can actually verify.

What We Declared on This Site, and What We Left Out

We added sameAs to this site's Organization schema on June 14, 2026, for a specific reason: a company we have no connection to operates under a nearly identical name on a different domain extension. The property exists for exactly that reason.

What we declared: our official Instagram profile, which we control and whose bio links back to mentionlab.ai (reciprocity, per the test above), plus a link to KomTop, our parent company, not a third party, which is why naming it here doesn't break this article's rule.

The list is short on purpose: every candidate has to pass all three questions, and a young site doesn't have ten defensible URLs waiting to be claimed. What we didn't declare: a Wikidata entry, since we don't have one yet.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "MentionLab",
  "url": "https://mentionlab.ai",
  "sameAs": [
    "https://www.instagram.com/mentionlab.ai",
    "https://www.komtop.fr"
  ]
}

Being honest about the result matters as much as the implementation: we can't isolate a measurable effect from one property change. sameAs isn't a switch you flip, it's one signal among many on a site that's still young, with no backlink profile yet. What we can observe fits into what AI visibility actually measures: a cleaner identity signal, not a ranking event.