This methodology describes how structured data should be planned, deployed and maintained so it actually improves AI retrievability rather than merely existing on the page.
Principle — entity-driven, not format-driven
Structured data is only valuable when it describes entities your market asks about. The methodology starts from your entity inventory, not from schema type lists. If a schema type does not map to an entity your buyers query, it is decorative.
Step 1 — Entity inventory
List the entities the site represents: the organisation, services, products, locations, team members, and the relationships between them. This inventory becomes the contract the markup must express.
Step 2 — Mapping to schema.org
Map each entity to its schema.org type (Organization, Service, Product, Place, Person, FAQPage where appropriate) and define the properties that must be populated. Where the entity has external identifiers (Wikidata, Crunchbase, company register), plan how they are referenced.
Step 3 — Implementation plan
Decide deployment mechanics: hand-authored JSON-LD blocks, a component-driven generator, or a combination. The site’s own approach — schema generated from content collections at build time — is the pattern to imitate: one source of truth, no drift between content and markup.
Step 4 — Validation
Validate every block for syntax, type correctness and property coverage using schema.org validation tooling plus a strict parser. Validation is part of the deployment pipeline, not a post-deployment chore.
Step 5 — Verification against engines
After deployment, verify how engines interpret the markup: entity descriptions in answers, knowledge-graph appearance, and structured-data diagnostics. The markup is only successful when it changes what engines understand.
Step 6 — Maintenance
Structured data decays: content changes, entities evolve, standards update. The methodology schedules quarterly audits against the entity inventory and re-verification after major schema.org updates. The structured data service runs this cycle; the entity optimisation service covers the entity-consistency work that complements markup.