The audit on https://aideazz.xyz scored 100/100 today. A yacht charter or luxury villa owner in Panama who asks why ChatGPT and Gemini rarely cite them receives the same score of zero on the checks that matter. Listing sites win because they pass every crawler test while most owner-run domains fail the first three.
How AI Assistants Choose Recommendations
Why Listing Sites Dominate Today
Booking.com, VRBO, and YachtWorld maintain dedicated teams that update robots.txt, add fresh llms.txt, and keep JSON-LD current for every property. Their domains have existed for decades, accumulate steady links, and serve pages under 1.5 MB that load in well under 1.5 s. An individual owner who lists only on those platforms inherits their visibility without doing the work. When a tourist asks Gemini for a luxury villa in Panama the model returns the listing site because the structured data explicitly marks the exact location, price range, and availability dates—facts the assistant can extract and cite. The owner-run site that forwards traffic to a GoDaddy domain or relies on a single Facebook page fails the sameAs and FAQPage tests and therefore never appears in the generated answer. The citation-probe.log that ran 6.5 h ago still shows only error objects; no owner domain has broken into the top results yet.
What an Owner-Run Site Must Do to Be Cited
Concrete Edits That Move the Score from 0 to 100
Place exactly one H1 at the top. Follow it with H2 headings that read as questions: “How many guests fit on a 50-foot catamaran in Panama?”, “What crew is included in the luxury villa rate near Bocas del Toro?”. Insert the answers as short paragraphs or bullet lists so the extractable-facts check passes. Add freshness dates to every page; the audit looks for any date signal. Declare canonical URL and Open Graph image so social and AI crawlers agree on the primary version. Run the visibility-audit.ts engine after each change; it returns the 0–100 score in four weighted groups. The commit history from the last 48 h shows 12 commits in aideazz and 12 in cto-aipa that implemented exactly these steps for EspaLuz country pages. The same workflow—question headings plus entity links—raises the Panama rental site from invisible to citable. The concierge-selftest.log that passed in 3436 ms confirms the pipeline can deliver results once the data is structured correctly.
Measuring Progress Without Guessing
The atlas-ga4-sync.log shows 0 atlas_ rows for the last three days, which means no traffic data is reaching the dashboard yet. The espaluz-funnel.log simply states “sent to Telegram,” confirming the lead hand-off works but not whether AI referrals have begun. The free audit remains the only reliable measurement: it runs 34 deterministic checks and returns a reproducible grade. Until that grade reaches 100 the owner cannot know whether ChatGPT or Gemini will cite the site. The audit today on aideazz.xyz reached 100/100 after the JSON-LD and llms.txt were added; the identical checklist applies to any luxury rental domain. The process dragontrade-dashboard has stayed online for 78 d with 68 MB memory; the same stability is possible for an owner site once the technical foundation checks pass.
Frequently Asked Questions
Q: How many checks must pass before the visibility audit returns 100/100?
A: All 34. The engine 1.2.0 scores four groups whose weights total 100 percent: 25 percent crawler access, 25 percent structured data, 30 percent answer-readiness, and 20 percent technical foundation. The site scored 100/100 today after the last 48 h of commits.
Q: Does adding prettier photos or a booking calendar improve the AI citation score?
A: No. The audit ignores images and JavaScript-rendered calendars. It only measures raw HTML content, JSON-LD, headings, and crawler headers. The technical-foundation group explicitly checks that content is present in raw HTML, not JS-only.
Q: Can I rely on a listing site alone and still get direct recommendations from ChatGPT?
A: Usually not. The model cites the listing domain, not the owner’s name or private contact. To appear with your own brand the owner-run page must pass the same 34 checks independently.
Q: What is the cost of the written diagnostic that maps each failed check to an edit?
A: The paid diagnostic is $100, credited if the client continues. It produces the report plus a 45-minute video call; the free audit at https://aideazz.xyz/api remains the first step.
— Elena Revicheva · AIdeazz AI Lab · Portfolio · Professional Outlook (PDF)
EspaLuz: live your new country in Spanish. Confidence, belonging and independence for international families in Latin America (web, WhatsApp, Telegram; 14 days free): aideazz.xyz/espaluz