/ Methodology
How we actually test this. No black box.
Real, research-validated traveler prompts, run against ChatGPT, Gemini, Perplexity, and Google AI Mode, in the languages your guests actually use.
/ What we track
Real traveler prompts, not "best hotel in [city]."
We test with research-validated traveler prompts, run against ChatGPT, Gemini, Perplexity, and Google AI Mode, in the languages your guests actually use.
Not generic category terms. Real phrasing like "boutique hotel near the medina with a pool" or "family-friendly resort with kids club." Full prompt transparency: you see every question, exactly as written.
/ What we verify, not infer
We don't guess at causality.
If review volume, Google Business Profile completeness, or a schema fix correlates with a visibility change, we say so only once we've tracked it. Not before.
/ A known limitation, stated plainly
Some AI answers pull from live web retrieval. Others come from trained memory.
We're actively building the ability to tell you which is which, because the fix is different for each. Until then, treat single-prompt snapshots as a signal, not gospel. That's why we test in volume, not one-off.
/ Minimum sample sizes
Before/after comparisons only show once we have a real signal.
Before/after comparisons only get shown to you once we've run enough prompts to be a real signal, not noise dressed up as a trend line.
/ GET STARTED
Find out who AI is recommending instead of you, before it becomes a habit travelers can't unlearn.
See the gap. Close the gap. Own your AI visibility.