Structured Data

What this parameter measures

Machine-readable structured data is content an AI system can parse directly, and this parameter checks whether your page publishes any. friendly4AI scans the page HTML for JSON-LD blocks (<script type="application/ld+json">), extracts each block's @type tokens, and checks them against a list of high-value schema.org types. Microdata can be merged in as an optional extractor, but it is gated by configuration and off by default, so the score is driven by JSON-LD. To stay safe on adversarial pages, the processor caps how many distinct types it records (SEC-6) and how large each JSON-LD block can be before it is parsed.

Why it matters for AI-readiness

Structured data tells AI engines exactly what your content is. It names whether a page describes an Organization, a Product, an Article, or a set of FAQ answers, instead of forcing them to infer it from prose. Schema markup powers rich results, feeds knowledge graphs, and makes your entities easier for ChatGPT, Gemini, and Perplexity to extract and cite accurately. Pages with clear @type declarations read as well-described and trustworthy, which raises the odds your facts are surfaced correctly.

How we score it

Under the v4.4 methodology, this Discovery parameter is scored in three tiers and reports only the id=3 structured_data behaviour. A page passes (100) when it carries at least one key schema.org type — Organization, WebSite, Product, Service, FAQPage, BreadcrumbList, Article, HowTo, VideoObject, or Event (matched case-insensitively) — or when two or more valid JSON-LD blocks are present, signalling rich coverage. It earns a partial (50) when at least one JSON-LD block exists but it contains none of those key types and there is only a single block — markup is present but limited. It fails (0) when no JSON-LD structured data is found at all. Malformed blocks are counted but do not by themselves raise the score.

How to fix common issues

• Add a JSON-LD block in a <script type="application/ld+json"> tag describing your primary entity (for example Organization and WebSite on the homepage).
• Use at least one key type from the list above; a single generic block without a key type only reaches Partial.
• Add a second valid block (such as BreadcrumbList or FAQPage) to reach the top tier when no single key type applies.
• Validate your markup with Google's Rich Results Test so malformed JSON-LD is fixed before it ships.
• Keep the @type accurate to the page content; mislabeled schema undermines trust and may be ignored.
• Re-scan after publishing to confirm the blocks parse and the key types are detected.