Entity mapping
An entity map is a working document that lists every entity your site needs to cover (your organization, people, products, places and topics) with its identifiers, key attributes, relationships and the one page that is its primary home. It turns entity research into a content, linking and schema plan, and becomes the source of truth for consistent facts.
What goes in an entity map?
| Column | Example |
|---|---|
| Entity name and aliases | Example Co; Example Company |
| Type | Organization / LocalBusiness subtype |
| Identifiers | Site @id, Wikidata Q-ID, Knowledge Graph ID |
| Key attributes | Founded 2009; HQ San Diego; phone |
| Relationships | founder → Jane Doe; serves → San Diego County |
| Primary page | /about |
| Supporting pages | /team, /locations/san-diego |
| Profiles (sameAs) | LinkedIn, Wikidata, industry directory |
| Status | Panel? Wikidata item? Facts consistent? |
The template is practitioner practice. Practitioner practice
How do you build one?
- Collect: gather entities using the finding entities methods.
- Identify: add Wikidata and Knowledge Graph IDs where they exist.[1][2]
- Classify: choose the most specific schema.org type for each.[3]
- Assign: give each core entity one primary page; avoid two pages competing for the same entity.
- Relate: write down the relationships you will express in text, links and markup.
- Standardise facts: agree the canonical name, address, dates and descriptions.
How does the map drive the site?
- Content: gaps show where a page is missing (topical authority).
- Internal links: every mention links to the entity’s primary page (guide to Internal linking for entities).
- Schema: each entity gets one
@id, reused site-wide (@id). - Off-site: the standard facts are copied to profiles and citations (NAP).
How often should you update it?
Whenever facts change (new address, new leader, new service) and at least at each audit. Out-of-date facts are a common source of wrong knowledge panels and AI answers. Practitioner practice
Which identifiers to record for each entity
An entity map earns its keep when it records the identifier each system uses, so that a mismatch can be found later. For each entity, consider these columns (the list of identifier forms is documented; the template is practitioner practice):
- Site
@id. The IRI your own JSON-LD uses for the entity; it should never change once published.[4] - Wikidata Q-ID or URI, if the entity has an item; items need to be notable and verifiable.[1]
- Knowledge Graph ID (
@idfrom the Search API), plus the type list returned.[5] - Registry codes for organizations:
iso6523Code(DUNS, GLN or LEI prefixes),naics,taxID.[6] - Official profile URLs that you will list in
sameAs.[7] - A one-line disambiguating description that you can reuse in visible text and
disambiguatingDescription.[8]
Frequently asked questions
Is an entity map the same as a keyword map?
No. A keyword map assigns queries to pages; an entity map assigns things, with identifiers and facts. Most sites benefit from both.
What tool should I use?
A spreadsheet is enough. The important part is one agreed source of truth for names, facts and primary pages.
See also
References
Pages accessed September 29, 2026 unless a date is given. See all sources and our editorial policy.
- ^ "Help: Items". Wikidata.
- ^ "Google Knowledge Graph Search API". Google for Developers.
- ^ "Data model". Schema.org.
- ^ "JSON-LD 1.1: A JSON-based Serialization for Linked Data". W3C Recommendation. Accessed October 29, 2026.
- ^ "Method: entities.search (Knowledge Graph Search API reference)". Google for Developers. Last updated June 26, 2024. Accessed October 29, 2026.
- ^ "Organization structured data". Google Search Central. Last updated September 8, 2026. Accessed October 29, 2026.
- ^ "sameAs property". Schema.org. Accessed October 29, 2026.
- ^ "Thing type". Schema.org. Accessed October 29, 2026.