Completeness without consistency produces a trust penalty. A domain with 12 Organization schema properties that conflict across pages is cited less accurately than a domain with 4 consistent properties. This week's data on the consistency gap and how to close it.
Structured data advice in AI visibility typically focuses on completeness — add more properties, cover more schema types, fill every optional field. The assumption is that more schema equals more citation. The data shows this is wrong when consistency is missing.
A domain with 12 Organization schema properties that contradict each other across pages — different names, mismatched founding dates, inconsistent sameAs URLs — is cited less accurately by LLMs than a domain with 4 properties that are perfectly consistent. The model detects the contradiction and applies a trust discount to the entire entity.
How entity trust penalties work
LLMs do not cite structured data naively. During inference, the model cross-references entity claims across sources: the Organization schema on your homepage, your Wikidata entity, your LinkedIn profile, your Crunchbase page, and your Wikipedia article. When these sources agree, entity confidence is high. When they contradict, confidence drops.
The most common contradictions in the tracked domain set:
Company name variants: "Veezow" on the website, "Veezow Inc." on LinkedIn, "Veezow Technologies" on Crunchbase. The model cannot resolve which is canonical. Citation accuracy drops — the entity may be cited under the wrong name or omitted when name-matching is a factor.
Website URL inconsistency: "https://veezow.com" on the homepage schema, "http://www.veezow.com" on Wikidata, no URL on LinkedIn. These are the same domain but the model treats them as unverified unless a sameAs link or canonical redirect explicitly connects them.
Founding date discrepancies: a common cause of citation errors in "when was [company] founded" queries. If Wikidata says 2023 and Crunchbase says 2022, the model cites the most-frequently-appearing value — which may be wrong.
The consistency audit
Running a consistency audit takes 30-45 minutes. For each of these fields, compare the value across your homepage schema, Wikidata, LinkedIn, and Crunchbase:
- Company name (exact capitalisation)
- Website URL (canonical form, including or excluding www and trailing slash)
- Founding date (year and month if known)
- Headquarters city and country
- CEO or founder name and title
- Industry/category label
Any discrepancy is a trust signal to fix. The fix is always: pick the canonical value, update the outlier sources to match, and verify through Google's Rich Results Test.
Completeness after consistency
Once your core entity properties are consistent across sources, completeness becomes the lever. The highest-impact additional properties by citation lift:
- sameAs links (each additional verified source increases entity confidence)
- description (natural-language description of the company, matching across sources)
- numberOfEmployees (a specific number, not a range where possible)
- areaServed or serviceArea (which markets or geographies you operate in)
- foundingDate (full ISO 8601 date, not just year)
Add these in order. Do not add them until core consistency is confirmed — a new sameAs link pointing to a source that contradicts your other data worsens the trust score.
This week's data
Across the domain set, entities with perfect consistency across 4+ sources had a mean citation accuracy score of 87%. Entities with at least one major contradiction across sources had a mean citation accuracy score of 61% — a 26-point gap.
Completeness alone (measured by property count) had a much weaker correlation with citation accuracy than consistency. The 10% of domains with the highest property count but at least one major contradiction scored below the median.
The lesson: audit consistency before adding properties. A lean, consistent entity schema outperforms a comprehensive but contradictory one.
Scan your domain to check your current structured data consistency score and see which properties are inconsistent across your entity sources.
Put this into practice
See how your domain scores on the signals covered in this edition. Veezow runs a free AI visibility scan — robots, sitemap, structured data, bot access, and off-site presence.
Run a free scan →