Two organizations can share a name. A homepage may also use a product name where a public reference page uses the parent company. Either case complicates an automated lookup, even when the site metadata is perfectly readable.
MachineRead's Wikipedia and Wikidata check searches for a small set of names and website references derived from the homepage. A match is worth inspecting. It is not enough, on its own, to establish that the returned record describes the same organization.
The lookup starts with homepage claims
MachineRead's Wikipedia & Wikidata Entity check begins with the fetched homepage. It derives up to four candidate names from og:site_name, JSON-LD Organization, LocalBusiness, or WebSite names, the page title prefix, and a domain-derived fallback. That candidate set is deliberately small.
The check then asks two public Wikimedia surfaces a narrow question. For Wikidata, it first looks for an exact official-website match against homepage URL variants. If that is not found, it searches labels with a limit of three results and accepts exact case-insensitive label equality. For Wikipedia, it searches page titles with a limit of three results and accepts exact title equality, including a parenthetical title whose base title exactly matches.
Those choices make the check deterministic enough to score inside a site audit. They also show why the result should not be overread. A bounded search can miss a relevant entity, return a same-name collision, or find a page that still needs human review.
Notability is a separate editorial question
Wikipedia's General Notability Guideline says a topic is presumed suitable for a stand-alone article or list when it has received significant coverage in reliable sources independent of the subject. It also says "presumed" means an assumption, not a guarantee, that the subject merits an article (Wikipedia, "Notability").
That policy is useful here because it cuts against a common shortcut. Notability is not the same thing as an API response. MachineRead does not evaluate source independence, coverage depth, article eligibility, Wikidata item quality, or editorial consensus. It only reports what its bounded lookup observed under the check's rules.
A positive result is a reason to inspect the matched page or item. A missing result can be normal for a small or young brand; it is not a judgment about the organization or the quality of its work.
Cached results are historical evidence
An entity finding may report a cached lookup rather than a fresh request to Wikipedia or Wikidata. Read the date attached to it. The external article or item can change after the stored observation.
If the result matters to a current decision, inspect the public record directly. Repeating the website audit during the cache window should not be mistaken for a new independent entity check.
Refusal and ambiguity are not absence
A refused or throttled lookup leaves the question unanswered. MachineRead reports an unknown-evidence warning for those cases rather than treating the refusal as proof that no entity exists. Read the state and finding, not just the points.
An ordinary no-match result also has limits. The relevant item may use another name, have a disambiguated title, lack a website property, or fall outside the small result set. Start by checking which public name your own homepage supplies, then search for the record directly before deciding it is absent.
How to use the finding
Treat the entity check as a triage signal, not a verdict. Start by reading the candidate names the homepage publishes. If those names are stale or inconsistent, fix the site metadata first. Then inspect any matched Wikipedia article or Wikidata item directly before making public claims from it.
If the check finds both Wikipedia and Wikidata evidence, verify that the matched records are actually about the same entity as the site and that their public facts are current. If it finds only one surface, look for a benign explanation before treating the gap as a problem. If it finds neither, decide whether that matters for the site's goals before attempting any outreach or documentation work.
Do not manufacture entity evidence for a score. Wikipedia and Wikidata have their own editorial and data-quality norms. Earned coverage, source quality, and public facts come first; audit output comes after.
Works Cited
Wikipedia. "Notability." Wikipedia, https://en.wikipedia.org/wiki/Wikipedia:Notability. Accessed 9 Sept. 2026.
See also
- Methodology reference - documents the Wikipedia & Wikidata Entity check group
- When a site uses three names for the same thing - separates page-published entity metadata from off-site verification
- Related: schema.org JSON-LD and search discovery