Most travelers don't wait for a phone line to open before asking a question anymore. Skift and Oracle Hospitality surveyed thousands of travelers and hotel executives in the spring of 2022. They wanted to know how comfortable people felt with automated service.
The answer leaned heavily toward yes: 77% said they were interested in using automated messaging or chatbots for customer service requests at hotels.
That figure predates the current wave of generative AI. It's only grown more relevant since travel and hospitality AI chatbot leads increasingly decide whether an inquiry turns into a booking or just disappears.
A missed call or an unanswered contact form doesn't just cost one booking. It costs the name, the travel dates and the reason to follow up, before a competitor's site collects the same thing faster.
Why Travel and Hospitality AI Chatbot Leads Aren't One-Size-Fits-All
Willingness to use one isn't even across a guest list. Younger travelers are far more open to it than older ones, and the two groups need different treatment from the same bot.
A bot tuned only for quick, transactional answers will frustrate an older guest who wanted a phone number. One with an easy handoff to a person, offered early rather than buried in a menu, serves both groups without picking a side.
Statista's tracking of AI use in hospitality and AI use in travel and tourism lays out how wide that gap actually is:
| What was measured | Figure | As of |
|---|---|---|
| US Millennials who'd support a brand using AI for customer service | ~70% | Statista tracking |
| US Baby Boomers who'd support the same | ~35% | Statista tracking |
| Consumers worldwide who'd used a mobile AI chatbot to plan a trip | 22% | October 2023 |
| Global hotel chains planning to implement AI agents | ~40% | 2025 survey |
| Travel and logistics companies whose AI agent use has passed the pilot stage | <10% | 2025 survey |
A chatbot built only for the confident half of that table will lose the rest of the room. One built to hand off smoothly catches both.
Hotel Chains Are Planning Faster Than They're Shipping

Photo by Vitaly Gariev on Pexels
Ambition is running ahead of deployment, based on that last row. Plenty of chains have announced plans. Far fewer have actually gotten an AI agent past a pilot.
That gap is an opening for a smaller operator. Whoever gets a chatbot handling real lead capture now, while most of the market is still planning, gets a head start before traveler expectations and the tooling catch up to each other.
Building that well looks more like AI and automation work than installing a plugin. It means training the bot on what a specific property or agency actually offers, and deciding exactly where it should stop and hand off to a person.
Skip that upkeep and a chatbot turns into the kind of project we've described in The Real Cost of AI at Work Is the Cleanup, where the mess shows up after launch, not before.
The properties furthest behind aren't the ones without a chatbot. They're the ones whose chatbot only answers questions and never captures anything, so every after-hours conversation ends the same way it did before the bot existed: nobody following up.
We've written before about what actually works in AI chatbots for customer service, and the short version is that routing matters more than the reply.
Large platforms aren't waiting either. Booking.com and Expedia both struck partnerships with OpenAI in late 2025 to put their inventory inside ChatGPT, per Statista's tracking of AI in travel and tourism. A guest can now start a trip search on a platform that isn't yours at all.
Capturing that guest the moment they land on your own site, rather than losing them to a slower reply, is the more winnable fight. It pairs naturally with the rest of a property's digital marketing and growth work rather than sitting apart from it.
Cover photo by Matheus Bertelli on Pexels
Sources
- Travelers Want High-Tech, Low-Touch Hotel Stays Shows New Oracle Survey — PR Newswire / Oracle Hospitality
- Artificial intelligence (AI) use in hospitality - statistics & facts — Statista





























