Why is AI search tracking different from SEO tracking?
AI search tracking is different because no tool equivalent to Google Search Console exists for it: Search Console shows exactly which queries triggered your listing and how often it was clicked, and AI search tools expose none of that.
ChatGPT, Perplexity, and Google AI Overviews do not expose impression or click data. The only way to measure your presence is to ask those tools the same questions your prospective guests would ask, and record the results.
This is manual, but it is accurate. And it produces data you can act on.
How do I set up my monthly query set?
Build a set of 8-12 queries that represent how real travellers describe your destination and property type. Run this exact set every month. Do not change the queries between months; consistency is what makes the data useful.
Query categories to cover
Destination-type queries (most important):
- “Best boutique hotels in [your city]”
- “Where to stay in [your city] for a long weekend”
- “Luxury hotels in [your neighbourhood or district]”
Feature-specific queries (tests your amenity signal):
- “Hotels in [your city] with [your key feature: spa, rooftop, pool, restaurant]”
- “Pet-friendly hotels in [your city]”
- “[Property type] near [local landmark]”
Direct brand queries (baseline check):
- “[Your exact hotel name]”
- “[Your hotel name] + [your city]”
Competitor context queries (measures relative position):
- “Best alternatives to [a well-known hotel in your market]“
Where to run queries
Run each query in:
- ChatGPT (GPT-4, no browsing, to test training data signal)
- Perplexity (tests real-time web indexing)
- Google AI Overviews (if visible in your region, tests integration with Google’s entity graph)
How do I record the results?
Use a simple spreadsheet with these columns:
| Date | Tool | Query | Named? | Position | Description used |
|---|---|---|---|---|---|
| 2026-08-01 | ChatGPT | ”boutique hotels Lausanne” | Yes | 2nd of 4 | ”lakeside views, independent” |
| 2026-08-01 | Perplexity | ”boutique hotels Lausanne” | No | (none) | (none) |
Run this monthly. After 3-4 months you will have enough data to see trends.
What does the data tell you?
The pattern across tools tells you where the gap is: a thin training-data signal, a weak live web presence, inconsistent entity data, or a foundational gap in all three.
Named in ChatGPT but not Perplexity: Your training-data signal is working, but your live web presence is weak. Focus on website indexation and external links.
Named in Perplexity but not ChatGPT: Your live web presence is solid, but your training-data signal is thin. Focus on editorial coverage and platform consistency.
Named with wrong description: Your entity data is inconsistent across sources. Run a hotel AI visibility audit to find the inconsistencies.
Not named anywhere: You have a foundational gap, usually schema, NAP consistency, or editorial presence. See why is my hotel not in ChatGPT.
What supplementary signals should I watch?
Two supplementary signals add context beyond query testing: referral traffic from AI tools, and the language guests use when asked how they found you.
Referral traffic from AI tools. In Google Analytics 4, check the Source/Medium report for traffic from chatgpt.com, perplexity.ai, and bing.com/chat. An upward trend in these referrers correlates with improving AI visibility.
Direct booking inquiry language. Ask guests how they found you. If you start hearing “ChatGPT recommended you” or “I asked Perplexity,” that is qualitative confirmation your presence is growing.
See what the AI presence score measures and how long hotel AI optimisation takes.
See also: How to report AI search visibility to hotel owners