Schema.org
Schema.org assesses how clearly machines can understand a page's entities and content through structured data. It contributes 10% to the GEO score.
What is assessed?
The audit extracts JSON-LD and Microdata from the HTML. The score considers:
- available and valid markup
- high-value types such as
Organization,LocalBusiness,WebSite,Article,Product,FAQPage,HowTo,PersonorBreadcrumbList - multiple types used in a meaningful combination
sameAslinks to verified profiles- links to Wikipedia or Wikidata
Invalid JSON-LD reduces the score. Markup without a type relevant to the page content receives only partial credit.
How to improve the score
- Choose the type that accurately describes the visible page content.
- Complete required and recommended properties.
- Use stable IDs so entities can be recognized across pages.
- Link only verified official profiles through
sameAs. - Validate JSON-LD after every change and avoid conflicting information.
Structured data should always reflect visible content. Additional claims that cannot be verified on the page do not improve actual trustworthiness.