Guests now ask ChatGPT, Gemini, Claude and Perplexity where to book. Watch it happen live — then track whether your hotels make the shortlist, who gets named instead, and whether the guest is sent to your site or an OTA.
ChatGPT, Gemini, Claude and Perplexity draw on different data and answer in their own way. The same guest question can put you at #2 on one engine and leave you invisible on the next.
Same hotel, same question, same week — in the answer on 1 engine out of 4. That's why one manual check tells you nothing.
Visibility per query, per engine, per market — the share of checks where your hotel makes the answer. Miss the shortlist and there's no booking to win.
→ Visibility & core coverageYour rank when shown, and which hotels the AI puts ahead of you. We track who appears instead of you and who appears alongside.
→ Rank & share of voiceThe link in the answer decides who gets the money: your own site — a full-rate direct booking — or an OTA, with commission out the door.
→ Destination: direct vs OTAThe sources feeding the answer — OTAs, review sites, or your own pages. If the engines never cite your site, the fix starts there.
→ Citations & sourcesEvery check is a real question asked from a real guest's country, captured as a timestamped screenshot of the assistant's answer. You see exactly what the AI said, when, and where.
ChatGPTDE
ClaudeLV
PerplexityFR
GeminiGBA dashboard built for a hotelier, not a data analyst: how visible you are, what's pulling it down, who's beating you, and what to do next.
| Hotel | Health | Shown | Avg rank | SoV | Status |
|---|---|---|---|---|---|
| Riverside Bay Hotel | 53 | 45% | #3.8 | 14% | Not visible |
| Riverside Central Hotel | 59 | 47% | #3.4 | 23% | OTA-cited |
| Harbour View Hotel | 63 | 88% | #2.8 | 31% | OTA-cited |
| Old Town Hotel | 67 | 100% | #3.1 | 27% | Outranked |
| Garden Court | 71 | 99% | #3.9 | 18% | Outranked |
| Pier 9 Hotel | 74 | 73% | #2.9 | 25% | OTA-cited |
| Lakeside Lodge | 56 | 45% | #4.1 | 12% | Not visible |
| Meridian Plaza | 58 | 88% | #3.2 | 22% | OTA-cited |
It audits each hotel's website for AI citability — can the engines' crawlers find it, read it and trust it enough to cite it — and turns every gap into a fix plan your web agency can deploy the same week.
Monthly, automatic. Crawler access, structured data, answer-shaped content, entity signals — your site the way AI engines actually read it.
Every finding is a P1–P3 item: what's broken, why it costs you citations, how to fix it — and the minutes it takes.
The webmaster pack: a brief plus ready-to-paste files. Forward one email to your web agency.
The next audit checks items off automatically, reopens anything that regresses — and your citation trend shows whether it worked.
Recommendations follow current best practices for AI search — they improve your odds of being cited, but how each engine ranks and answers stays outside anyone's control.
Every guest an AI sends to your own site instead of an OTA is a full-rate booking with zero commission. Priced per hotel, per month.
Annual billing −15%; from 10 hotels −10% — discounts stack. Pairs with Rate Radar as the Direct-Booking Defense Suite. Prices per hotel, per month, excl. VAT.
Provisional pricing during early access — final plans confirmed before onboarding.
The four that matter most for travel today — ChatGPT, Gemini, Claude and Perplexity. Each is checked independently, because they draw on different data and often give different answers.
No. SEO is about ranking blue links on a results page. This is about whether an AI names your hotel in its answer and where it sends the guest. Different signals, different playbook — and increasingly where the booking decision happens.
We don't rely on one snapshot. Each query is checked repeatedly from a neutral session in the guest's own country, captured as a screenshot, and rolled into a trend. You see the pattern over time — with the evidence behind every number.
When an assistant recommends you, it often includes a link. We classify where that link goes: Direct (your own site — a full-rate booking), OTA (Booking, Expedia — commission out the door), or Other. Being recommended but routed to an OTA is a hidden leak we surface.
No — that would only prove the AI has heard of you. We ask the way a real guest does: "best spa hotels in Riga," "family hotels in Tallinn with breakfast." A 70/20/10 mix of high-intent, competitive and broad queries tests whether you get recommended unprompted.
The fix side of the product. It audits each hotel's website for AI citability — whether AI crawlers can find, read and trust your site enough to cite it — and turns every gap into a prioritized fix plan for your webmaster, with ready-to-use files (robots.txt diff, structured data, llms.txt). After your team deploys, we re-audit and verify the fix automatically.
No. Give us your hotels and target markets and you're live in days — no PMS, no channel-manager integration, no work for your IT team.
EU-hosted and GDPR-ready, with each client's data isolated. The data we collect is public AI-assistant answers about hotels — not personal data. (Early-access terms apply.)
A note on honesty: our recommendations follow current best practices for AI search — they improve your odds of being cited, but how each engine ultimately ranks and answers stays outside anyone's control.
Tell us your hotels and target markets and we'll run a first read across all four engines — with the screenshots to prove it. Early access is open now.
ip@signs.lv