Agentic AI is reshaping how travelers find and book hotels
By Cathleen Draper
Artificial intelligence is no longer just a buzzword in hospitality – it’s actively running operations, fielding guest inquiries, and filling hotel rooms. At the center of this transformation is a new class of AI: Agentic.
Agentic AI plans, takes action, and completes tasks without constant supervision. You simply give it a goal. Because it’s autonomous, rather than spitting out simple text or images, this type of AI builds upon the generative models we’re now familiar with.
AI agents show up in a myriad of ways across operations. Certain solutions monitor inventory and predict procurement needs. Voice AI agents handle reservations and recommend amenities or local attractions to guests. AI-driven revenue management tools analyze booking trends and competitor rates to optimize yours. Agents review and respond to RFPs for room blocks and events.
But agentic AI isn’t just reshaping hotel operations. These agents are revolutionizing hotel bookings and distribution.
The New World of Search
More and more, travelers are turning to AI platforms to discover, compare, and book hotels. Instead of browsing OTAs or Google search results, they’re leveraging services like ChatGPT, Gemini, and Perplexity, which pull options, assess availability and pricing, and even make reservations, with some circumventing third parties and booking direct.
International Data Corporation, a global market intelligence and data provider, predicts that by 2030, 30 percent of travel bookings will be executed by AI agents.
Major brands are jumping on board. In May, Wyndham launched a native ChatGPT app, allowing users to explore some 8,400 hotels across Wyndham’s portfolio, filtering by location and amenities and refining results through prompts. When they’re ready to book, the app sends them directly to Wyndham’s booking site.
What’s driving this shift is customer preference.
“They don’t want to stay in the same cookie-cutter hotel. They want unique experiences,” said Romal (RJ) Jayswal, co-founder and CEO of StayNow, an AI-powered travel planning app, and Predictive Minds, an AI-powered revenue and reputation management platform.
Guests also want a stay that fits their exact wants and needs. Say a person is traveling to Dallas for a conference. Before agentic AI, they’d simply search for hotels in downtown Dallas under a certain price point.
Now, they want more. For example, Jayswal explained, that same traveler today will ask AI to “find a hotel in downtown Dallas with a rooftop bar and vegetarian food options. AI will first do the research, [then show] the hotel which is close by, and it has a rooftop bar, and whatever other preferences, and it will confirm and book it.”
With agentic AI, guests are “no longer forced down this typical path of ‘tell me your dates and how many adults, and then I’ll show 100 hotels that you can pick from,’” said Ashwin Kamlani, CEO of Hyperfunnel, which is developing an interactive voice-driven AI travel agent for smart TVs. “It should just give you the three best options.”
As traditional, static searches become less common, hotels need to shake up their strategy.
A standard FAQ on a hotel’s website is not enough to appear in agentic-driven results. Hoteliers need to go beyond check-in and check-out times, when breakfast is served, and whether their property has a pool.
“AI chatbots are looking for more intent base, or why are people even going there,” Jayswal said. Take the example of a guest seeking a hotel in downtown Dallas during a conference. Maybe they want a hotel that keeps them in the thick of the conference buzz, or maybe they’d like a quieter stay away from the crowds.
“That intent and the nuance of why a guest chooses your hotel, that is very important,” Jayswal said. “The general questions about your location, things to do, and your amenities – all of that is standard. It’s 50 percent, but the other 50 percent is your true differentiator. Out of all the 10 [recommended] hotels, why will somebody choose yours?”
How to Get Discovered
A hotel’s offerings must be detailed and machine-readable. Leveraging SEO to simply rank in Google search is out the door.
“We almost have to retrain our minds about what information is important and what we should make publicly available,” Kamlani said. “Now, with AI, people can ask anything. And they should be able to find the answer.”
For example, a hotel historically might have never listed how many handicapped parking spots are available, and exactly how many are on the right side of the property. Now, Kamlani said, hoteliers should put that level of detail somewhere on their site so it shows up if a potential guest asks that question in ChatGPT. Some hotels might even upload their blueprints so AI agents can answer anything about the layout.
“That’s what the future is about,” Kamlani said. “How do you arm these new language learning models (LLMs) with as much data as you possibly can? Because it’s no longer about usability. It’s about access to every last detail.”
To arm these LLMs on the front-end, web content must be machine-readable. Use conversational language and a natural, relaxed tone that mirrors how people speak and ask questions. Follow a question-and-answer format. Ensure descriptions are as detailed as possible – yes, you might already note that pets are allowed, but also include that there is a pet fee, your hotel features a dog run outside, or you provide a water bowl in the room.
Not all of this information has to appear directly on your website. Rather, hoteliers can code these answers as structured data, also called schema markup. This data in your website’s HTML doesn’t show up for users browsing your page, but it helps AI models understand content and display these extra details in the results they pull.
In this data, hoteliers can describe all the spaces and amenities their hotel offers and list local attractions and nearby dining options. This might be where they list the number of handicapped parking spaces and where they’re located, or the blueprint of their property.
One thing hotels should avoid: Pretending to be something they’re not.
“If you’re a three-star hotel, and you’re not the most beautiful option, but you’re going after the customer that’s looking to save money, then you should just say that,” Kamlani said. “Because what AI is doing very well is it’s not just taking at face value.”
“You might be excellent at SEO, and 10 years ago, you’d build a website positioning yourself as a luxury hotel, even though you’re not,” he continued. “But you can’t get away with that anymore.”
AI leverages what the world says about a hotel, not just what a hotel’s website says about it. Reviews – and any public information about your property on the Internet – play a major role in discoverability. AI agents can read every single one in the matter of minutes. Hoteliers should encourage guests to leave reviews, and thoughtfully and clearly respond to each one.
“The best thing you can do is to just be authentic and real about what you’re selling,” Kamlani said. “You can’t possibly control what the world is saying about you. You can’t go into Reddit and manipulate what’s in there to impact what ChatGPT pulls back as a result for the consumer.”
For owners that feel they’re on an equal footing with every other comparable property in the area, Jayswal recommends they need to dig deep to find what differentiates their hotel from the rest.
“Everybody is unique,” he said. Maybe their property has a busy pool, or there’s a social element to their hotel. Or they have a volleyball court or are close to a trail ideal for a morning walk.
“That is the superpower of AI – that AI is searching for that thing, and if you’re able to mention that, then your property will come up in the chatbots, not the OTA or Google searches,” Jayswal said.
The Fight for Guest Relationships
Optimizing for AI discoverability is essential in today’s marketplace since agents can increasingly drive direct bookings. For hotels positioned to appear in an agent’s search, less and less of their profits could be taken by OTAs.
More and more companies are becoming middlemen between a hotel and its guests, too. At the end of April, Uber introduced a new vertical: Hotel bookings. In partnership with Expedia, Uber will now offer more than 700,000 hotel booking options for U.S. travelers.
That’s bad news for hoteliers, Jayswal said, because it adds another layer, one more middleman taking a piece of the revenue pie. “That is what, as a hotelier, should be bothering you,” he continued. “Because the more layers that come in, the piece of pie is getting smaller, little bit by little bit. It’s just adding more and more o top of who is actually delivering the value.”
Kamlani disagrees. “Super apps are coming to the US. Both Chase and Capital One are now OTAs.” He argues that these apps will open fresh opportunities for the industry, with new, more profitable, and more transparent distribution models.
Though they differ on the rise of these super apps, Jayswal and Kamlani, and their companies building AI travel agents, are unified in their goal: Connecting hoteliers directly to their guests.
If there’s one thing hoteliers must understand, Jayswal said, it’s that they need to begin adopting AI solutions and optimizing for AI discoverability now.
“Otherwise, they are going to become housekeepers,” he said. “You’re just cleaning rooms, getting things ready, and letting all of these guys own the guest. And whoever owns the guest is in control.
“They need to have a strategy to own the guest, not the asset. Owning an asset is a lost business. Owning the guest is the real business, and they need to get into that side of the business as quickly as possible.”
Image: SB Creative/stock.adobe.com
Leave a Reply