Does your site serve structured data?
It checks for schema coverage, business identity and breadcrumb markup.
What Check Engine measures
Pages intended for search are checked for JSON-LD, microdata and recognised script-built types. No detected markup produces high severity. Missing identity and identity built only by scripts get different explanations. At least 5 indexable pages without breadcrumb markup produce a low finding. Coverage above zero but below 0.4 produces another low finding. With no indexable pages, coverage stays unknown. For example, markup coverage of exactly 0.4 stays outside the partial-coverage finding.
The check reads structured types in server responses and types known to be built by scripts. Objects it cannot read remain unknown. Review coverage before treating an apparent gap as a reason to add duplicate markup.
How to read the result
No schema and absent identity can be high, script-only identity medium, and breadcrumb or coverage gaps low. Presence of a type is not proof that all required properties exist or that a particular rich result is available.
Why it matters
Coverage and correctness are separate questions. Knowing where markup appears helps locate a template gap, while the type and completeness checks still need to establish whether the declarations fit the content.
How to fix it
Choose markup that accurately describes the business and page. Add it to the appropriate shared template and serve important identity in readable HTML where practical. Keep breadcrumb declarations aligned with the visible route. Check several page types after editing, then validate values and ownership instead of stopping at a presence result.
How to fix structured data errors
What we found in field audits
Problems in Structured data were found on 11 of the 14 sites we audited. These are area totals. Read the field-audit method.
Method note
backend/seo_engine/checks/structured.py, missing_schema, revision 0bb8e8d3.
Common questions
Can missing breadcrumbs be reported on a small schema sample?
The count must reach 5 indexable pages lacking breadcrumbs. Smaller sets do not reach that finding.
Does markup coverage of 0.4 trigger the partial-coverage finding?
No. The partial-coverage comparison applies below 0.4 and above zero. Equality stays outside that result.
Is identity markup built only by scripts treated like served identity?
No. The report rates script-only identity medium, with a separate explanation for absence. Served JSON-LD and microdata are also recognised.