AI Visibility in Hospitality: What 3,600 AI Responses and 1,337 Website Audits Reveal

How AI models recommend accommodation is no longer a future scenario. Information-seeking queries already account for nearly 50% of all ChatGPT conversations. There are no blue links. The AI either recommends your property or it doesn't.

Between August 2025 and February 2026, Nokumo Research conducted two original studies to find out what determines whether AI recommends your property — and whether your website is ready for the answer.

Selected Findings

Booking.com appeared in 95.3% of all queries tested — capturing 1 in 7 of all AI-cited URLs.

Only 5.7% of accommodation websites were detected by any AI model. The other 94.3% are invisible.

The language you search in changes what AI recommends. Croatian-language queries produce up to 10 percentage points more OTA dependency than German — even for Austrian hotels.

What separates AI-visible properties from invisible ones isn't what the industry expects. Clean URLs and trust signals have 2–3x the impact of schema markup.

77.1% of accommodation websites have no booking engine visible to AI. The average property operates at 38.1 out of 100 on digital readiness.

The full report covers the complete dataset, country-specific breakdowns, the intent × provider matrix, a self-assessment scorecard, a 90-day action plan, and expert contributions from four industry specialists.

Download the full report — 42 pages, free

Contributors

Dario Alfirević (Nokumo) - Chief Researcher and Editor-in-Chief

Tihana Alfirević (Nokumo) - Editor

Niclas Aunin (ALLMO.ai) — AI Visibility Measurement

Sophia Milewski (Smartness / SmartPricing) — Dynamic Pricing as AI Trust Signal

Domagoj Davidović (MinusMinus Agency) — Branding in the Age of AI

Arne Petersen (VRM Days) — The German Market Gap

In the Media:

Frequently Asked Questions

How does AI decide which hotels to recommend?

AI models like ChatGPT, Gemini, and Perplexity don't rank hotels the way Google does. They recommend three to five specific properties by name — and everything else is invisible. Our study of 450 queries across four AI models found that AI citations are driven by structured data coverage (Booking.com appears in 95.3% of queries), third-party authority (TripAdvisor, DMO listings, editorial coverage), and website quality — especially clean URLs, trust signals, and factual content. The full breakdown, including which factors matter most and which matter less than expected, is in the 42-page report.

What percentage of hotels are invisible to AI search?

94.3%. Our audit of 1,337 accommodation websites across Germany, Austria, Switzerland, Slovenia, and Croatia found that only 76 properties (5.7%) were detected in any AI model response. The average digital readiness score is just 38.1 out of 100, with 77.1% of properties lacking a booking engine and 60.8% having no schema markup at all. The report includes a self-assessment scorecard so you can benchmark your own property.

Which AI model is best for hotel direct bookings?

Does search language affect which hotels AI recommends?

What does the report include?

42 pages covering: the complete dataset from 3,600 AI responses and 1,337 website audits, country-specific breakdowns for Germany, Austria, Switzerland, Slovenia and Croatia, a full intent × provider matrix, a digital readiness self-assessment scorecard, a prioritized 90-day action plan with technical and non-technical tracks, and expert contributions from four industry specialists on AI visibility measurement, dynamic pricing, branding, and market structure.

Who is this report for?

Hotel operators and marketing managers who want to understand whether AI can find their property. DMOs and tourism boards evaluating their ecosystem's digital readiness. Hospitality technology providers building products for the AI search era. Revenue managers exploring how pricing strategy intersects with AI visibility. The data covers independent hotels, hotel chains, vacation rentals, and hostels across five European markets.