TL;DR
- DirectBooker has signed supply deals with five of the world’s ten largest hotel groups, including BWH Hotels and Radisson
- Live integrations now exist as a ChatGPT app and a Claude connector, surfacing rates, availability and rooms inside conversational AI
- The set-up uses Model Context Protocol (MCP) to pipe live hotel data into the assistants, then redirects to the hotel’s own website to close the booking
- Coverage extends to Cendyn, eviivo, Mirai, SiteMinder and Roiback, pulling in thousands of independent and boutique properties
- Hotels currently pay OTAs 15-25% commission; this is the industry’s first serious move to win that traffic back through AI
What’s Happening at DirectBooker
DirectBooker quietly went live this month with something the hotel sector has been waiting twenty years for. The platform now connects five of the world’s ten largest hotel groups, including BWH Hotels and Radisson Hotel Group, directly into ChatGPT and Claude. Loyalty programmes covering more than 250 million members are now surfaceable inside the same conversational assistants your clients are starting to plan their holidays in.
The clever bit sits underneath. DirectBooker has structured its data around Model Context Protocol, the open standard that lets AI assistants pull live information from outside sources. Rates, availability, room types, commercial terms, all of it now flows into ChatGPT and Claude in real time, with the booking itself completed on the hotel’s official website. Independents are in the mix too, through tie-ups with Cendyn, eviivo, Mirai, SiteMinder and Roiback.
The Practical Bit for Your Bookings
This matters because the route your client takes to a hotel booking is changing in front of you. For two decades, the journey was OTA, hotel website, or an agent. Now there is a fourth path: a conversational AI assistant that knows live rates, knows the loyalty perks, and hands the client off to the hotel’s own checkout. Sanjay Vakil, DirectBooker’s chief executive and co-founder, framed it bluntly. Hotels have been on the defensive for twenty years, watching OTAs take the top-of-funnel customer and charging 15 to 25 percent for the privilege.
For your role as the trusted human in the loop, this is genuinely useful intel. Clients who experiment with AI-planned trips will increasingly hit hotel options surfaced through these tools. The conversation you can have with them is not AI versus agent, but AI for browsing, agent for the booking that protects them when it counts.
What to Tell Clients About AI Hotel Search
If a client mentions they have been asking ChatGPT about hotels, that is a green light, not a threat. AI-assisted browsing is brilliant for shortlisting. It is far less brilliant at handling a date change, a complex family rate, a loyalty-tier upgrade, or the moment something goes wrong at 11pm on a Saturday. You stay the person who picks up the phone, knows the property, and has the supplier relationship to fix things.
A useful framing for the client conversation: ChatGPT can show them ten Mauritius hotels with current rates in thirty seconds. What it cannot do is tell them which of those ten has the best beach for a toddler, which one had a refurb six months ago, or which one will hold a room if their flight is cancelled. That is your work, and it is where the value of the relationship lives. The clients who use AI to research will arrive at you better informed and clearer on what they want, not less likely to book through you.
Where Agents Win From Here
The hotels making this move are openly trying to reclaim the customer relationship. That means richer rates, better perks for direct bookings, and a stronger story for you to take to clients about why an agent-supported direct rate beats whatever ChatGPT initially suggested.
