Start with what is published
For a company, the name, official website address and related genuine profiles may be appropriate. For an article, the heading, author, date and image may be appropriate if they are actually present. A software product description should match its current capabilities and terms.
Google emphasizes that markup must match visible information and the rules for the specific type. Do not add reviews, ratings or prices that are absent from the page and cannot be verified. A large number of types does not automatically make the content better. Sometimes a smaller but consistent set of fields is enough.
Syntax, meaning and eligibility for display
JSON validation establishes whether a record can be parsed. Required-property validation relates to the selected type and scenario. Factual accuracy is another level. Even a correct record does not guarantee a rich result in search, much less a brand recommendation in ChatGPT.
For example, a service is in early access and its form only saves a preferred plan. Markup should not turn that into an already active paid subscription. First clarify the actual offer, then align the public text, price and machine-readable fields. The same applies to product availability and authorship.
Check both the template and real instances
Inspect several page types: the homepage, an article and a service page. A common problem is a shared Organization or Article template that references someone else’s name, an old domain or a nonexistent author. After a change, check the published HTML, not just the file in the editor.
Use the official validator for the relevant search scenario and review the meaning manually. If you add a date, it should reflect a real event. If a sameAs link points to a profile, make sure it belongs to the same organization rather than simply a similar name.
How AURA GEO helps
The AURA audit records the presence and basic correctness of JSON-LD, and the agent can prepare a supported edit based on project or page facts. This is not full certification against every schema.org rule. Deployment requires authorized access to the client’s hosting environment.
After applying the change, save it along with verification results. Then measure real answers and visits separately: technical care matters, but it does not replace the usefulness of the content itself.
Conditions and limitations
- AURA’s basic check does not cover every search rich result rule.
- The presence of JSON-LD is no basis for promising a recommendation.
Explore the details
The markup agrees with the content and is checked after publication.
Prepare company information