What Is the sameAs Property in Schema Markup and Why Does It Matter for AI?

The sameAs property is schema's cross-platform entity linker, the machine-readable equivalent of a business telling Google and every AI system reading its site, "we are the same business across all of these platforms." It matters because without it, systems that should be combining scattered information about a business into one picture may instead treat each platform as a separate, unrelated entity.

People search for this concept under several spellings — sameAs schema, sameAs schema markup, schema sameAs, and, often typed with a space, "same as schema" since it reads that way out loud even though the property name itself has no space in it. All of them point at the same JSON-LD property, covered below.

How sameAs actually works

"sameAs": [
  "https://www.practo.com/bengaluru/clinic/sharma-skin-clinic",
  "https://www.facebook.com/SharmaSkinClinic",
  "https://www.linkedin.com/company/sharma-skin-clinic",
  "https://www.justdial.com/Bengaluru/Sharma-Skin-Clinic"
]

When this list appears in a business's LocalBusiness schema, an AI system reading it knows to combine information from every listed source into a single entity profile, reviews from Practo, contact details from JustDial, company information from LinkedIn, all attributed to one business rather than treated as coincidentally similar names. The property itself is defined at sameas, and it works alongside the broader knowledge-graph concept it feeds into.

sameAs schema markup in full context

The snippet above shows the property in isolation, but sameAs schema markup normally sits inside a complete LocalBusiness block, not on its own. A fuller example, for the same clinic:

{
  "@context": "https://schema.org",
  "@type": "LocalBusiness",
  "name": "Sharma Skin Clinic",
  "url": "https://sharmaskinclinic.example.com",
  "telephone": "+91-80-XXXX-XXXX",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "12 MG Road",
    "addressLocality": "Bengaluru",
    "addressRegion": "KA",
    "postalCode": "560001",
    "addressCountry": "IN"
  },
  "sameAs": [
    "https://www.practo.com/bengaluru/clinic/sharma-skin-clinic",
    "https://www.facebook.com/SharmaSkinClinic",
    "https://www.linkedin.com/company/sharma-skin-clinic",
    "https://www.justdial.com/Bengaluru/Sharma-Skin-Clinic",
    "https://www.google.com/maps/place/?q=place_id:EXAMPLE_PLACE_ID"
  ]
}

This is what makes schema sameAs functional rather than decorative: it lives inside the same JSON-LD block as the business's core NAP data, so an AI system reading the page gets the address, phone number, and the cross-platform profile list together, as one entity record rather than fragments to reconcile separately.

What happens without it

Without sameAs, an AI system may treat a Google Business Profile, a Practo listing, and a JustDial entry as three separate entities that happen to share a name, especially if there's any variation in how the name appears across them. That fragmentation directly reduces AI citation confidence, since the system has no explicit signal telling it these are one business rather than three loosely related ones. Entity-disambiguation covers this problem from the AI system's side of the equation.

What actually belongs in a sameAs list

The primary vertical directory profile is worth including first, Practo for healthcare, 99acres for real estate, Zomato for restaurants, whichever platform is the authoritative one for that industry. A Facebook Business Page, a LinkedIn Company Page, a JustDial listing, and the Google Maps listing itself round out the standard set. A Wikipedia page, if one exists for a large enough brand, and any other genuinely high-authority profile can be added too. Wikidata entries are worth including where they exist, since AI systems weight structured knowledge-base entries heavily.

What shouldn't go in the list

Low-authority, unverified, or spam directories don't belong here, even if a listing technically exists on them. Including only platforms where the profile is accurate and actively maintained matters more than maximizing the count of URLs in the array, since a sameAs list pointing to a stale or wrong profile actively creates the same ambiguity the property is meant to solve.

Keeping this current

A sameAs list is only as good as its last update. New verified profiles should get added as they go live, and any URL that's been deleted, renamed, or replaced needs pruning, a maintenance habit covered more broadly at keep-schema-updated-business-changes. Angryturtle's citations-nap module tracks the directory profiles a business actually holds, which makes keeping a sameAs list accurate a byproduct of citation management rather than a separate manual task.

FAQ

Does sameAs help local pack ranking directly? No, similarly to other schema properties. Schema-affect-local-pack-rank covers why schema's value sits mostly in entity clarity and AI Overview eligibility rather than local pack position.

How many URLs should a sameAs list realistically include? There's no fixed number. Four to eight accurate, active profile URLs across the platforms that genuinely matter for the business's industry is a reasonable range, more than that starts diluting rather than reinforcing.

Should sameAs include competitor comparison pages or directory listing pages that mention the business? No. sameAs is for the business's own verified profiles on other platforms, not third-party pages that merely reference it.

Is "same as schema" the same thing as sameAs schema markup? Yes. "Same as schema," written as two words, is simply how people type out the sameAs property when they're describing it in plain English rather than quoting the JSON key itself. The property, the markup, and the concept referred to as "same as schema" are all the same thing shown in the JSON-LD examples above.