Entity trust
How to Fix Broken sameAs Links Without Creating a False Entity Graph
A fail-closed workflow for auditing Organization identity, validating sameAs targets, removing false equivalence, and monitoring entity links after deployment.

In brief
- sameAs asserts that a target URL identifies the exact same entity.
- Use one canonical Organization node, validate every target, and remove related-but-not-identical pages.
- Test rendered templates and monitor target ownership, redirects, and duplicate emitters over time.
Sections in this article
TL;DR
- Use sameAs only for pages that identify the exact same entity-not partners, articles, directories, or merely related profiles.
- Validate ownership, entity name, canonical target, status, visibility, and cross-profile consistency before publishing each URL.
- Keep one stable Organization @id and one accountable emitter of the entity graph across templates and plugins.
Key takeaways
sameAs means exact identity, not relevance, partnership, or mention.
Maintain one canonical entity record and one stable Organization @id.
Validate ownership, canonical target, and visible identity for every URL.
Consolidate duplicate emitters instead of stacking another JSON-LD block.
Monitor semantic identity changes, not only HTTP availability.
chema.org defines sameAs as a URL that unambiguously identifies the same item. That is a stronger statement than “related to,” “mentioned by,” or “we have a page there.” A partner profile, customer review, press article, reseller page, or topical Wikipedia article is not automatically the same entity.
Treat each sameAs value as an edge in an identity graph. The source node is your Organization or Person. The target must resolve to a page whose visible content and ownership identify that exact entity. If the target is ambiguous, inaccessible, repurposed, or controlled by somebody else, remove the edge until it can be verified.
Google's structured-data documentation explains how machine-readable markup helps describe a page. Independent implementation references and webmaster forums add practical context about how sites assemble Organization graphs. Your own validation still decides whether a target is true for your entity.
Identity rule
If you cannot finish the sentence “this URL identifies the same legal or public entity as our Organization node,” do not put it in sameAs.
In this article
- 1.What sameAs actually asserts
- 2.Building the identity inventory
- 3.Validating every target
- 4.Repairing the JSON-LD graph
- 5.Testing rendered output
- 6.Monitoring identity drift
Start with one canonical entity record: official name, commonly used brand name, legal name when public and relevant, canonical website URL, stable Organization @id, logo, contact point, and approved public profiles. Separate current profiles from historical names, product brands, regional subsidiaries, founders, and partner organizations.
Record where each identity claim is emitted. Common emitters include the theme, SEO plugin, ecommerce app, tag manager, page builder, custom component, and server-rendered template. Duplicate emitters can create several Organization nodes with different names and sameAs arrays.
Do not merge different entities for convenience. A parent company, subsidiary, product, and founder may be related, but they need explicit relationships such as parentOrganization, brand, or founder where appropriate-not a false sameAs edge.
Identity inventory fields that make repairs auditable.
| Field | Example purpose | Owner |
|---|---|---|
| Canonical @id | Stable graph identifier on the site | Web/platform owner |
| Public name set | Recognized names and spelling | Brand owner |
| Approved profiles | Exact-entity public pages | Brand/comms owner |
| Emitter map | Template, plugin, app, component | Engineering owner |
| Last verified | Target and identity review time | Named reviewer |
Resolve the URL safely and record the final canonical target. Confirm a successful response, public visibility, correct entity name, recognizable branding, and a plausible ownership relationship. A redirect to a generic login page, marketplace search, unrelated successor account, or another entity fails the check.
Compare identity details across targets. The exact address or name format may differ, but unexplained conflicts in company name, website, region, or ownership need review. Do not use one profile to “prove” another in a circular chain; retain the evidence you used to approve each target.
Webmaster and Shopify community discussions are valuable for discovering duplicate-emitter problems. They do not establish your ownership. Verification should come from the live target, account control, and an internal approved-profile registry.
Checklist
- URL resolves to a public, canonical page rather than search or login
- Visible page identifies the exact same entity as the source node
- Profile links back to the official domain when the platform supports it
- Name, logo, region, and ownership do not contradict the canonical record
- An owner and last-verified timestamp are stored
See where your brand appears in AI answers - and where it does not.
Run a first-party brand check across supported answer engines. Results are measured without a promised citation or conversion. Browse all free tools
Choose one accountable component to emit the primary Organization node. Give it a stable HTTPS @id anchored to the canonical domain, and reference that same ID wherever the organization appears. Remove duplicate primary nodes or connect legitimate related nodes explicitly.
Keep the sameAs array small and verified. More links do not create a stronger entity. A short set of exact, maintained public identities is safer than a long directory containing abandoned accounts, press coverage, or related brands.
When an SEO plugin already emits a valid Organization graph, configure or extend that owner rather than injecting a second script. Yoast documents one implementation pattern; other stacks differ. The governing requirement is one coherent graph, not one preferred vendor.
Repair decisions for common URLs.
| Candidate URL | Relationship | sameAs? |
|---|---|---|
| Official company profile | Exact same entity | Yes, after validation |
| Founder profile | Person related to organization | No; model the person relationship |
| Press article | Mentions the entity | No |
| Subsidiary website | Related legal entity | No; use organization relationship |
| Abandoned social handle | Previously controlled | Remove until recovered |
Fetch the final server response and inspect the rendered DOM because client-side apps and plugins can add markup after the template is generated. Enumerate every Organization and Person node, every @id, and every sameAs value. Then compare the result with the approved entity inventory.
Test representative templates: home, about, product, article, author, and localized pages. A global header can emit the graph everywhere, while an article plugin may add a conflicting publisher node only on posts. The audit should report the page template and emitter responsible for every duplicate.
After deployment, re-fetch the live page and save a before/after graph diff. A structured-data testing tool can help inspect syntax, but semantic identity remains a human-governed assertion. Passing a parser does not make a false equivalence true.
- Fetch server HTML and render the page with client scripts enabled
- Extract every JSON-LD node, @id, relationship, and sameAs URL
- Map each node to its template, plugin, or application emitter
- Compare exact targets with the approved identity registry
- Deploy, re-fetch, and store the graph-level before/after diff
Syntax is not identity proof
A validator can confirm valid JSON-LD syntax. It cannot confirm that two public pages represent the same real-world entity.
Revalidate sameAs targets on a schedule and whenever a brand, domain, profile, plugin, or theme changes. Watch HTTP status, final URL, canonical target, public visibility, displayed entity name, and the graph emitted by representative pages.
Use reasoned states: verified, changed, inaccessible, ambiguous, and removed. Do not silently preserve a broken edge because the URL still returns 200. A profile can resolve successfully while now representing another organization.
The repair artifact should list the old edge, evidence checked, decision, new graph, and verifier. This gives brand and engineering owners a shared record and prevents a future plugin update from reintroducing the same false identity assertion.
FAQ
Should a press article be placed in sameAs?
No. It mentions or describes the entity but does not identify the article itself as the same organization.
Can a subsidiary use the parent company URL in sameAs?
Not merely because they are related. Model the organizational relationship explicitly and reserve sameAs for exact identity.
How many sameAs links should an Organization have?
There is no useful target count. Include only exact, public, owned or verified identities that the organization can maintain.
Does a structured-data validator prove the links are correct?
It can validate syntax and recognized vocabulary. It cannot prove real-world ownership or exact entity equivalence.
References and further reading
These links are provided for direct inspection. A reference is not treated as proof of every statement in this article.
- 1.Schema.org sameAsschema.org
- 2.Google introduction to structured datadevelopers.google.com
- 3.Wikidata introductionwikidata.org
- 4.Yoast Organization schema implementation referencedeveloper.yoast.com
- 5.Webmasters discussion of Organization markupwebmasters.stackexchange.com
Written by
EdenRank Editorial Team
The product and editorial team documents repeatable ways to inspect AI-answer visibility, source evidence, and content operations.
Expertise
Want insights like this for your own brand?
Talk to the teamKeep building the topical graph.
AI Visibility Monitoring Pricing in 2026: What Teams Actually Pay
Compare AI visibility monitoring costs, run volume, evidence exports, and exclusions before choosing a plan or approving a vendor quote.
The 30-Day ChatGPT Brand Visibility Fix: From Invisible to Cited
Most brands are invisible in ChatGPT not because of poor content, but because they fail the three filters ChatGPT uses to cite sources. Here's how to pass all three in 30 days.
How to Use Review Data to Improve AI Search Citations
Use review data with valid Review markup, FTC-compliant collection, dated platform evidence, frozen prompts, and separate outcome measures.