Structured Data

Structured data states facts in defined fields so machines do not have to infer them from prose. A LocalBusiness object can include name, address, phone, and hours; an Article can identify its author and dates; and a Dataset can describe published research.

For a local service business, common types include LocalBusiness or a more specific subtype (Dentist, Plumber, LegalService) with name, address, phone, hours, and service area; Service for offered services; Article with author and dates for editorial content; BreadcrumbList for site structure; and Organization with sameAs links to corroborating profiles. Add only types supported by visible content and a defined use.

Markup describes content but does not improve the underlying page. Structured data types are also not direct ranking bonuses. Google retired FAQ rich results in May 2026, so FAQ markup no longer earns that special display.

Pages that answered useful questions retained that information after the FAQ display changed. The markup only described those answers. Build the page for readers first, then add structured data that accurately represents its visible content.

Validate the rendered markup after deployment and after template changes. Check that URLs, dates, prices, ratings, and identifiers match the visible page and canonical source. A valid syntax check confirms the format, but it does not verify that the claim is true or eligible for a search feature.

For AI systems, structured data can provide unambiguous facts such as hours, address, and services. Keep the markup accurate, limited to supported claims, and consistent with visible content. See JSON-LD for an implementation approach that keeps both versions in sync.