Build an entity graph for your business by publishing a schema.org Organization node with a stable @id and a sameAs array linking to authoritative external profiles — Wikipedia, Wikidata, LinkedIn, Crunchbase, official social accounts — then referencing that same @id from every Article, QAPage, and Person node on the site.
Google's original Knowledge Graph announcement introduced the "things, not strings" framing that still drives entity resolution today: search engines and answer engines identify unambiguous entities and then attach content to them. The sameAs property is documented at schema.org/sameAs as the vocabulary link between your on-site entity and its off-site profiles.
A minimal working graph declares one Organization with @id, name, url, logo, and sameAs; a WebSite with its own @id and a publisher reference to the Organization; and, for each contributor, a Person with @id, name, jobTitle, and its own sameAs. Reuse the @id values everywhere — every Article's publisher and author point back to them.
Wikidata is the highest-leverage target: it is the machine-readable backbone that other knowledge graphs (Google's, Bing's, and the LLMs' pretraining corpora) draw on. A Wikidata Q-number in sameAs is worth more than three social links.
Validate the graph in the Rich Results Test and confirm all @id references resolve inside a single JSON-LD block.