This lesson closes the remediation module by showing how structured data states who and what an entity is — and how one wrong default explained three findings from earlier modules. It then shows the system we built to fix this site’s structured data and keep it fixed.
What this lesson covers
- What structured data can and cannot do
- The properties that define an entity
- Checking schema against content and the graph
- A real contradiction and its repair
- Verification in three layers
- A registry-driven system, measured before and after
In the audit snapshot, the site’s brand node was declared as both a Person and an Organization, with no sameAs links — a plugin default nobody had revisited. The lesson shows how that one node explains why the content is read as a person, why the brand had no clean Knowledge Graph match and why nothing corroborated it. The repair follows four steps: decide what the entity is, declare it once, point at independent proof and verify. The before and after are real. Since 26 October 2026 the site has an Organization with its own identifier, four official profiles and a founder link to a Person on her own domain.
The last part shows the system behind that fix: a six-step method of detect, classify, map, register, generate and validate, applied to the main site and the academy. The baseline audit found twelve issues on a site with zero validator errors, including 134 paid lessons with no markup and 92 posts with two identifiers for the same author. The release replaced them with one JSON-LD block per page, valid paywall markup and one identifier per person. Fixing the markup also uncovered 81 broken links on live pages, because the markup reads the same fields the page shows.
The lesson is careful about what the evidence supports. It quotes Google’s guidelines directly, notes that a valid result in the Rich Results Test does not guarantee any rich result, and presents a vendor’s controlled sameAs test with its small sample and short window stated alongside the numbers.
This is the last lesson of Module 5 in The AI-powered Entity Audit, followed by a practical that builds a registry and Claude-drafted JSON-LD. It suits anyone who manages a site’s structured data and wants it to agree with everything else the site says.