What you get from an entity audit
You get a written report with six parts: an entity inventory of every name and identifier in use, a schema graph review with corrected JSON-LD for your key templates, a sameAs and profile map, a Wikidata eligibility assessment, the AI assistant and search recognition findings with their sources, and a prioritised fix list.
| Deliverable | What it contains |
|---|---|
| Entity inventory | For each entity in scope: the canonical name, legitimate alternate names, variants to retire, the canonical page and the @id to use. |
| Schema graph review | What your current JSON-LD says, template by template; errors, conflicts and dangling references; and corrected, connected JSON-LD for the home page, about page, location pages and author pages. |
| sameAs and profile map | The profiles that should appear in sameAs, the ones that shouldn't, and inconsistencies in name, address, phone, website or description across them. |
| Wikidata assessment | Whether an item exists and is correct, whether the business meets Wikidata's notability policy, and the references that support it, or a plain "not yet" with the reason. |
| AI and search recognition findings | The questions asked, how each assistant or search engine answered, what it got wrong, and the sources those answers drew on. |
| Prioritised fix list | Every change, ranked, with who makes it (developer, content, profile owner) and exactly what to change. |
What the audit doesn't include
- Guaranteed outcomes. No one can promise how a search engine or AI assistant will describe a business. The audit removes the contradictions and gaps that cause mistakes; how and when systems pick up the changes is up to them.
- Creating profiles or Wikidata items on your behalf where the business doesn't qualify.
- Fake reviews, ratings or citations, or markup that describes things not on your pages.
After the audit
Most fixes are small: a name made consistent, a sameAs list corrected, one Organization @id used everywhere. Your team can make them from the fix list, using the free schema.technology generator for the markup. For the concepts behind them, the Ontology Courses lessons take about five minutes each.
Want an entity audit? Book a 30-minute call with Neural Command, or email [email protected].