It fetches your page, extracts every JSON-LD block, and for each recognized schema type compares the properties present against the required and recommended properties Google documents for that type. It then reports a completeness percentage per block, listing which properties you have and which you are missing, so you can see how fully each piece of structured data is filled out.

It scores the most common rich-result types: Article, Product, LocalBusiness, FAQPage, Organization, Recipe, and Event. For each it knows the required fields Google needs and the recommended fields that improve eligibility and display. Unrecognized types are still detected and listed, but only the supported types receive a detailed property breakdown.

For each schema block, the score weights required properties more heavily than recommended ones, because missing a required field can disqualify the rich result entirely. The percentage reflects how many of the expected properties for that type are present. A high percentage means the markup is thorough; a low one points to gaps that limit how richly Google can display it.

No. Validity asks whether the JSON-LD is syntactically correct and meets minimum requirements. Completeness asks how thoroughly you filled in the recommended properties that improve rich-result quality. A schema can be valid yet incomplete — for example a Product with a name and price but no image, brand, or aggregateRating. Use our JSON-LD Validator for validity and this tool for depth.

Google can only display the rich-result features your data supports. A Recipe with cook time, ratings, and nutrition can show a far richer result than one with just a name. Recommended properties also strengthen entity understanding. Filling them in raises your chance of enhanced listings and gives search and AI systems a clearer picture of your content.

It reads the HTML returned by the server, so it captures JSON-LD present in the initial response. Structured data injected later by client-side JavaScript may not appear. Most sites output schema server-side, but if yours is rendered in the browser you may need to check the rendered DOM separately. The tool tells you how many blocks it found so you can spot a mismatch.

Treat required missing properties as urgent, since they can block the rich result, and add recommended ones where they genuinely apply to your content. Do not invent data to fill fields — only add properties that are true for the page. After updating your markup, re-run the scorer and confirm with Google's Rich Results Test.