An operator cannot guarantee that an AI assistant will recommend a brand, but can audit the evidence available for discovery: crawl access, consistent entity facts, supported structured data, authentic reviews, and documented independent corroboration. The goal is a correct, inspectable evidence trail—not a speculative ranking score.
Last updated: July 24, 2026
Key Takeaways
- Crawl eligibility is necessary for search inclusion but does not guarantee placement or recommendation.
- Keep the brand name, canonical URL, description, logo, contacts, and policies consistent across first-party properties.
- Use structured data to describe supported facts, not to manufacture proof.
- Inventory authentic reviews and independent mentions without inventing a required count.
- Record the evidence source, owner, last verification date, and contradiction status.
Can documented search crawlers access the site?
Begin with eligibility. Check robots controls, page accessibility, canonical URLs, and whether important evidence is available without an unsupported interaction. OpenAI’s ChatGPT Search documentation says sites should not block OAI-SearchBot to be included, while also stating there is no way to guarantee top placement.
This distinction matters. Access enables consideration; it does not prove authority, quality, or recommendation. Track crawler access as one audit line rather than presenting it as a growth tactic.
Are the brand facts consistent across first-party properties?
Build a fact sheet for the brand’s preferred name, URL, description, logo, contact information, product names, founders or owners where public, and key policy pages. Compare the website, marketplace profiles, social profiles, directories, and structured data. Flag contradictions and assign an owner to resolve them.
Google’s Organization structured-data documentation says markup can help Google understand and disambiguate administrative details, with supported properties such as name, URL, logo, contact details, and sameAs. Markup should match visible, supported facts; it is not a quality certificate.
| Evidence class | Audit field | Operator action |
|---|---|---|
| Crawl access | Allowed crawler, accessible URL, canonical | Remove accidental blocks |
| Entity facts | Name, URL, description, logo, contact | Resolve contradictions |
| Claim support | Source, date, scope, direct evidence | Remove or qualify weak claims |
| Reviews | Platform, product, authenticity, disclosure | Preserve representative evidence |
| Independent mentions | Publisher, URL, claim corroborated | Document rather than imitate |
Which brand claims have direct evidence?
Separate claims supported by a product page, policy, specification, demonstration, case record, or independent source from claims repeated only by the brand. Record the precise wording each source supports. If the source is outdated or narrower than the claim, correct the claim rather than stretching the evidence.
DGP’s account of appearing in ChatGPT results reports a first-party observation that specific content and off-site mentions preceded citations. It does not establish a universal factor or causal formula. DGP’s AEO guide and GEO guide address content structure; this audit addresses evidence control.
Are customer reviews authentic and representative?
Inventory review sources, product association, dates, material connections, incentives, and moderation practices. Do not create, buy, rewrite, or selectively suppress reviews to produce a desired sentiment.
The FTC’s reviews and testimonials rule Q&A discusses fake or false reviews, sentiment-conditioned incentives, undisclosed insider relationships, and review suppression. Review governance belongs in the evidence audit because distorted evidence creates both trust and compliance risk.
What independent sources corroborate the brand?
List direct sources that independently describe the brand’s category, products, operating history, or reputation. Record what each source actually supports and when it was checked. A repeated press release is not the same as independent corroboration.
Finish with an evidence register: claim or fact, source URL, evidence class, owner, last verified date, contradiction, and next action. Recheck it after a brand change, site migration, product launch, or material policy update.
Frequently Asked Questions
Can structured data guarantee an AI recommendation?
No. Structured data can help a supported system understand stated facts, but it does not guarantee inclusion, ranking, quality judgment, or recommendation.
Does allowing OAI-SearchBot guarantee ChatGPT visibility?
No. OpenAI documents crawler access as necessary for inclusion while stating that top placement cannot be guaranteed.
What belongs in a brand evidence audit?
Include crawl access, consistent entity facts, structured data, claim support, authentic reviews, independent mentions, source URLs, owners, dates, and contradictions.
How often should brand evidence be reviewed?
Review it on a recurring cadence and after changes to the brand, domain, products, policies, profiles, structured data, or important third-party coverage.
Turn scattered proof into an evidence register: Get From Zero to Sovereign.
Follow the forthcoming book at The Social Proof Moat.