Generative AI is moving from an experimental travel tool into a meaningful part of accommodation research. Deloitte’s travel surveys show expected use of generative AI for trip planning rising from 8% of respondents in 2023 to 16% in 2024 and 24% in 2025. Among those planning to use it, more than half (54%) of intended users prioritized accommodation research.
This growth changes how hotels enter a traveller’s consideration set. Search engines and online travel agencies generally begin with standard fields such as destination, dates, price, star category and amenities. An AI tool on the other hand can process a more detailed request: a hotel with connecting rooms, a children’s club for a particular age group, gluten-free breakfast and an airport transfer after midnight.
The property may offer every one of those services. However, it will be difficult for an AI system to recommend the hotel when the information is vague, buried in a PDF, missing from the website or inconsistent across external listings. A hotel can therefore be available and suitable for the trip but remain absent from the shortlist.
From SEO Rankings to AI Recommendations
This creates a distinction between search engine optimisation, or SEO, and generative engine optimisation, or GEO. SEO helps a hotel rank in conventional search results and attract website visits. GEO focuses on whether AI systems can understand, verify and recommend the property in response to a specific traveller query. A hotel may perform well in SEO while remaining weakly represented in AI-generated answers.
For years, hotel content has relied on broad descriptions such as family-friendly, pet-friendly, accessible, sustainable and conveniently located. While these labels work as marketing signals and category tags, they do not provide enough detail to answer specific traveller requirements. Family-friendly, for example, does not show whether connecting rooms can be guaranteed, which ages the children’s club accepts, whether babysitting is available or if cots must be requested in advance. Pet-friendly does not explain size restrictions, charges or permitted areas. Similarly, close to the airport does not provide the expected journey time, shuttle schedule or hours of operation.
As AI-led discovery develops, each of these details becomes part of hotel distribution. Room information needs to include occupancy limits, bed configurations, connecting-room rules, dimensions, accessibility adaptations and extra-bed availability. Facility information should cover opening hours, age restrictions, dietary capabilities and advance-booking requirements. Location data needs measured distances, realistic travel times and transport options.
The requirement is becoming particularly important for meetings and events. A hotel may promote flexible event spaces, but an AI tool assessing a corporate brief needs room capacities under different seating layouts, ceiling heights, breakout inventory, audiovisual facilities, loading access, menus and sustainability certifications. If those fields cannot be identified, the property may be excluded before its sales team receives a request for proposal.
Hotel Groups Are Building for AI Discovery
IHG provides the clearest indication that hotel groups now view content architecture as part of distribution. The company is developing a fully AI-compatible platform that will structure highly specific, modular data for every property across its system of more than 6,900 hotels. IHG directly links the investment to natural-language search, personalised results and top-of-funnel visibility, with a phased rollout beginning in 2026, indicating the move of granular property data from a website-content function into a system-wide guest-acquisition capability.
Marriott has followed a similar path with Ask Bonvoy. The tool interprets natural-language queries and searches the company’s portfolio using verified Marriott-owned information, including details on restaurants, spas and golf facilities. Meanwhile, Accor and G6 Hospitality have extended hotel discovery into ChatGPT itself. ALL Accor allows travellers to search properties, availability and rates through a conversational interface, while the Motel 6 plugin lets users compare hotels, prices, amenities, photographs and availability before proceeding to book.
These initiatives connect property descriptions with live commercial information. An AI system becomes more useful when it can establish that a hotel has the requested feature, determine whether it is available for the required dates and direct the traveller towards a valid booking option.
Despite these capabilities, AI has not yet replaced traditional discovery and booking channels. Expedia Group’s 2026 research shows that only 8% of surveyed US and UK travellers relied on AI chatbots or agents for trip planning, compared to 59% using search engines and 49% using OTAs. While 53% felt comfortable receiving AI-generated suggestions, 68% still preferred booking through a trusted travel brand. Therefore, the defining commercial opportunity sits at the top of the search funnel. Travellers may continue to transact through hotel websites, OTAs, and established platforms, but AI will increasingly influence which properties make their shortlist.
AI Visibility Requires a Single Source of Truth
Data accuracy across every touchpoint is essential. AI engines continuously pull from hotel sites, OTAs, review platforms, and event databases. When key details like room counts or sustainability claims conflict, AI tools struggle to establish a single source of truth, directly risking missing out on high-intent recommendations and bookings.
Hotels consequently need a verified first-party source of truth and a process for keeping external listings aligned with it. Marketing teams can improve how attributes are presented, but operations must confirm facilities and policies. Sales must maintain event specifications, distribution teams must connect attributes with availability, and technology teams must keep the information structured and accessible.
Ultimately, AI search is turning hotel attributes into operational distribution infrastructure. A competitively priced property can still lose visibility when its product cannot be accurately understood or matched to a traveller’s request. Hotels that manage attributes as verified and consistently distributed data will be better placed to enter the shortlist before the booking journey moves elsewhere.