JSON-LD

JSON-LD (JSON for Linking Data) is one of three formats for embedding Schema.org structured data in web pages, alongside Microdata and RDFa. Google, Bing, and most AI assistants treat JSON-LD as the canonical format because it separates data from presentation: the structured data lives in a self-contained <script> tag and is invisible to users.

A typical JSON-LD block looks like a single JSON object with @context (always "https://schema.org"), @type (the Schema.org type, e.g., "LocalBusiness"), and the properties for that type. Multiple types can be combined in a single page using @graph notation, where the graph property holds an array of entities each with their own @type and @id for cross-reference.

A common pitfall: emitting JSON-LD via Next.js's metadata.other field produces a meta tag (<meta name="application/ld+json" content="..."/>), which Google and AI assistants do NOT parse. JSON-LD must live inside a script tag with type="application/ld+json" — typically rendered server-side via dangerouslySetInnerHTML in a React framework, or directly in HTML in non-React contexts.

Validate JSON-LD at search.google.com/test/rich-results before deployment. Google's rich-results validator catches schema-type mismatches, missing required properties, and malformed JSON.