This insight summarizes Blake Reiter's June 2026 Hotel News Resource article "The AI Recommendation is the New Battleground for Hotel Distribution", drawn from Lighthouse research presented at Luminate 2026. The study ran 4,545 ChatGPT prompts across 9 destinations and 5 traveler personas, generating 49,707 hotel mentions — and the patterns are not random.
Invisibility is the default state. Of thousands of available properties, only 2,721 unique hotels appeared at all. Coverage rates are brutally low in mature markets: Tokyo 10%, Paris 13%, and in Park City 33% of properties are effectively invisible to ChatGPT. The top 100 hotels globally captured 13% of all mentions — a power-law distribution, not a uniform one.
The bias is measurable. Four- and five-star properties dominate generic searches; three-star properties are nearly absent. Even for budget personas, recommendations skew luxury — 83% of business-travel and 73% of family-travel picks are 4–5 star. Branded chains crowd out independents: Marriott alone captures 27% of branded U.S. recommendations. Independents fare differently by market — Indianapolis (6.5%), Tokyo (10.5%), Brussels (28%).
Where ChatGPT pulls its answers from. 82% of cited sources are OTAs/metasearch and editorial/media sites. That means rate-parity, OTA listing hygiene, and earned media are now upstream of AI visibility. PR is no longer brand work — it is distribution.
Five operator actions Reiter recommends. (1) Treat PR as a distribution channel. (2) Audit OTA listings for staleness and inconsistency. (3) Rewrite content language to match the personas you want to attract — "play-led" descriptors (arts, attractions, nightlife) win generic searches. (4) Manually prompt ChatGPT in your markets to see where you stand. (5) Optimize the direct-booking funnel — AI-referred traffic tends to land on the property site, not the OTA.
"Your content is no longer just marketing copy. It's algorithm input." — Blake Reiter