JSON-LD

JSON-LD (JSON for Linked Data, RFC 8259) is a way to embed semantic / structured data in JSON format. On the web, it's used to give search engines a machine-readable description of page content -- enabling rich-result features (knowledge panels, FAQ accordions, breadcrumb trails, recipe cards, product price + rating displays).

The on-page form looks like:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "How to bake bread",
  "image": ["https://example.com/bread.jpg"],
  "datePublished": "2026-05-09",
  "author": {"@type": "Person", "name": "Jane Doe"}
}
</script>

Two key tags:

• @context -- the vocabulary URI. For the web, this is almost always https://schema.org (sometimes http://schema.org for legacy reasons; both work).
• @type -- the schema type. Common ones: WebSite, Organization, Article / NewsArticle, Product, BreadcrumbList, FAQPage, HowTo, Recipe, Event, LocalBusiness, Person.

Why JSON-LD vs other formats:

• Microdata (itemtype="..." attributes inline) and RDFa are alternatives but require sprinkling attributes throughout HTML markup -- changes to schema force changes to the page template.
• JSON-LD lives in a separate <script> block, so the schema can be generated server-side or by a CMS plugin without touching the HTML body. Google explicitly prefers JSON-LD over microdata for this reason.

Common bugs:

• Missing or wrong @context (must be https://schema.org).
• Missing @type.
• BreadcrumbList without numeric position on each ListItem (silently drops the rich result).
• Multiple Organization blocks confusing the publisher identity.
• Type-specific required-property gaps (Article without headline, Product without offers, etc.) -- BeaverCheck's per-type validator catches these.

Validation: Google's Rich Results Test is the canonical tool. Schema.org also publishes Schema Markup Validator for type-level conformance.