Today, you have to juggle several tools:
It is a case of all these tools (PMS, CRS, CRM, Q-Data) gathered into a single, smart, well-organized, and easily accessible box.
Conversational agents or platforms (Gemini, ChatGPT, Perplexity, etc.) will no longer need to search everywhere. They will open the box and access structured, reliable, real-time data directly, and thus they will be able to give answers or even complete a booking instantly.
The query could be: “Do you have availability next weekend for a soundproof room with USB-C outlets, and do you offer a vegetarian menu at your restaurant?”
If your MCP contains both your booking engine and a highly documented structured database like Q-Data, the AI agents will be able to handle complex requests like this. AI agents will no longer simply consult static data — the MCPs will allow the agents to access the most recent, up-to-date, precise information directly.
However, we are still in the early stages; there’s no certainty yet about when or how public LLMs (ChatGPT, Perplexity, Gemini, etc.) will connect directly to these hotel MCPs. But sooner or later, it will happen — for several reasons:
Through the MCP Registry, the central directory of the MCPs. It lists, verifies, and easily connects all compatible AI servers and tools. It is the key component that turns each MCP into an open, interoperable standard for AI governance and connectivity.
In the meantime, the most intelligent thing to do right now is to optimize your website and its ecosystem to ensure that it is readable in this world of generative search (SEO and AEO, FAQs, OTA updates, Google profile). Please see the previous article “Is SEO Dead? The Zero-Click Paradox.”
Possible Scenarios
The OTA Scenario
ChatGPT recently announced the launch of the Agentic Commerce Protocol (ACP), paving the way for direct e-commerce – and soon, for hotel bookings, without visible intermediaries. In this new ecosystem, OTAs appear as natural partners for generative AIs: their deep catalogues and smooth technical integration make them ideal providers for LLMs.
The result: OTAs could secure the monopoly on generative search, leading to biased search results with favourable offers depending on the commercial agreement of the hotel.
Moving forward the risk could be significant for the hotel industry, but for now, Booking.com and Expedia are the only players whose apps are directly integrated into ChatGPT’s search and monopoly is questionable To be continued…
The MCP Scenario
If LLMs and conversational agents gain direct access to hotel MCPs (or at least the MCPs of hotels which are equipped with one), they’ll be able to view real-time rates and a huge volume of richer data: amenities, services, guest reviews, photos etc. Travelers would then get a wider, more detailed and objective overview of all available options.
The Likely Reality
As usual, the truth will most likely lie somewhere in between. Ultimately, customers will be able to choose between going through an OTA or booking directly via the hotel’s MCP.
Hotels will therefore need to complement their official website with a proper MCP containing availability, rates, and of course data – lots of data (Q-Data).
See article: “What’s the Future of My Website?”
The hotels of tomorrow, will have not just one internal AI, but dozens of AIs:
An AI in the TV, an AI in the room-service robot, one at reception, one on the website, on the phone, in Teams, Slack, Copilot, etc.
Each of these AIs will most likely be installed by different providers, each building its own (more or less effective) set of data.
The result is inevitable: these AIs will be inconsistent with one another. On a daily basis it is likely that if you ask the same question to the TV AI and the website AI, you may get totally different answers.
Once again, it is the MCP which stands out as the one true source of truth, provided it’s based on a deep, detailed, well-structured database such as Q-Data.
Conversely, an unstructured base – such as a simple RAG-type mix of hotel sources (website, brochures, leaflets, etc.) – would only add confusion.
Moreover, maintaining and updating such a patchwork unstructured database would quickly become totally unmanageable.
See the June 2024 article (still relevant): “You Shall Suffer from Poor Data Quality.”
Hotels will need to rapidly make the transition from the official website, to the official FAQ, to the official MCP containing the official data (Q-Data).
The fact that you’ve read this far is already an excellent first step.
The next step is to carefully start to prepare the setup of your MCP server. Contact your PMS, CRM, booking engine, and AI solution providers to learn about their respective roadmaps. Then, get them all to collaborate with the aim of a coordinated rollout.
But before this, you can be effective right now:
-Optimize your website for generative search by integrating your FAQs – you need dynamic, multilingual FAQs, which are supported by a solid base of customer reviews.
Please see : “SEO and AEO with AI and Big Data.”
-Keep your Google Business profile and your data on all OTAs and Metasearch platforms up to date – data must be consistent everywhere to be credible and recognized as such by LLMs. For this, a tool like Q-Channel will be extremely useful.
Now ask yourself: What are the essential elements needed to choose and book a hotel?
The revolution is now and hotels must prepare for this real-time data revolution, built on verified structured information (Q-Data), within AI platforms.
When AI interfaces become bookable, the historical dominance of the OTAs may falter – opening up new conversational booking opportunities directly for hotels.
And all of this will rely on the MCPs.
We will therefore move from the official website to the official FAQ, and then to the official MCP containing the official data (Q-Data).
Amen!
OTAs are no longer fighting for clicks – they’re fighting for the conversation between the user and the AI agents
To remain visible, hotels must therefore inject their data directly into these agents.
Dynamic multilingual FAQs (Q-SEO) become a strategic asset: they provide LLMs with reliable answers and allow hotels to appear in conversational results – currently the only credible alternative to the OTA–LLM duopoly.
OTA integration relies on the Model Context Protocol (MCP) — the same technology already used by Q-Data.
The future of hotel visibility will depend on standardized, open, and inter-operable data flows:
availability, rates, and inventory (ARI), combined with structured hotel data (Q-Data).
And all of this will be listed in the MCP Registry.
We’ve been talking about this for 9 months. Day by day this is becoming the reality. For the future, development is likely to be very rapid.
And this is just the beginning — we haven’t even mentioned the A2A (Agent-to-Agent) Protocol, which is even more powerful.
It will allow a traveler’s agent to communicate directly with a hotel’s agent, discuss available services, negotiate upgrades, and confirm bookings — entirely autonomously.
See the article “How Personal Assistants Will Transform Hospitality.”
Hoteliers, you won’t be able to say you didn’t know, because we told you so…
Amen !
An MCP for hotels (Model Context Protocol) is a new open standard that connects all a hotel’s systems — PMS, CRS, CRM, and Q-Data — into one unified, intelligent platform. It centralizes availability, rates, descriptions, photos, and guest data in real time. This allows conversational public LLMs like ChatGPT, Gemini, or Perplexity and private LLMs like chatbot, AI phone, AI assistant, robot, smart TV…. to access verified hotel information instantly — and even complete bookings directly. The MCP for hotels is therefore becoming the cornerstone of AI-driven hospitality.
The MCP for hotels is reshaping the hospitality landscape by enabling direct connections between hotels and AI-powered assistants. Instead of searching static websites, conversational agents will interact with live hotel data to answer complex questions and process bookings instantly. This evolution reduces dependency on OTAs, improves guest experience, and ensures that hotels remain visible and competitive in the age of generative AI and agentic AI.
Hotels should begin by structuring and synchronizing their key data — availability, rates, and inventory (ARI) — along with detailed hotel descriptions with fully stuctured data (Q-Data). Next, they should collaborate with their PMS, CRM, and booking engine providers to ensure future MCP compatibility.
While waiting for full integration, hotels can already improve visibility by:
Preparing for the MCP today means being ready for AI-native bookings tomorrow.
© Image: Shutterstock
Learn the 10 proven methods to increase direct bookings and reduce third-party dependency.
Your website gets traffic, but are they booking? Engage visitors instantly with Velma answer questions 24/7, and boost direct revenue.
Download the free checklist 
Convert more visitors today 
