What it checks
We look for script blocks of type application/ld+json in the HTML the server sends, parse each one as JSON, and count the schema.org types they declare. We do not run JavaScript, so blocks injected by a tag manager are not seen; neither would a crawler see them without rendering.
Why it matters for search engines
Google's structured data documentation recommends JSON-LD as the format for marking up pages. Valid markup makes your page eligible for rich results; a block that does not parse is ignored, so a single trailing comma silently removes that eligibility.
Why it matters for AI assistants
A JSON-LD block is the one place on a page where the type of thing (an Organization, an Article, a Product), its name and its key facts are stated unambiguously. Assistants that build answers from page content can read that directly instead of guessing from layout.
How to fix it
Add one JSON-LD block in the head: Organization or WebSite on the home page, and the page type (Article, Product, FAQPage) where one fits. Validate it with a JSON linter before publishing; a block must be pure JSON, with no comments or trailing commas.
<script type="application/ld+json">
{"@context": "https://schema.org", "@type": "Organization",
"name": "Example Agency", "url": "https://example.com/"}
</script>What quietwatch reports
ok when at least one block parses, listing the types found. warn when there is no JSON-LD, or when a block fails to parse (with the block number and the parser's reason). unknown when the page could not be fetched.
Sources
All twelve checks · Previous: Time to first byte · Next: Content without JavaScript