MCP – an acronym I hadn’t heard anyone in hospitality use until a few weeks ago, when Tom James, MD at Bill’s, dropped it at a Propel Multi-Club event. By his own admission, he’d only come across the term a few weeks earlier.
Despite its newness, MCP could become one of the biggest enablers of real AI value in hospitality. The reality is, right now, AI on its own doesn’t do much beyond basic task handling.
You probably won’t hear ‘MCP’ mentioned in many boardrooms yet, but over the next few years I think it will quietly become one of the most important drivers of AI adoption in hospitality. So here’s a breakdown of what MCP is and why it matters.
What is MCP, in simple terms?
MCP = Model Context Protocol. It’s essentially a way for AI tools to securely connect to external systems and use real business data in context.
In other words, MCP allows AI tools like Claude or ChatGPT to connect to the rest of your operation. Reservations, EPOS, CRM, scheduling, guest data and more.
Think of it as a bridge between AI and your operational tech stack. Without that bridge, AI only knows what you manually tell it. With it, AI can operate with real-time business context.
For example:
Without MCP, AI can help write a marketing email or generate ten variations of it. Useful, but aside from saving copywriting time it hasn’t really created much value.
With MCP, that same AI could pull live CRM segments, identify lapsed high-spend guests, cross-reference booking availability and deploy targeted campaigns based on previous visit behaviour.
Same AI. Completely different outcome.
Why this matters for hospitality
Hospitality businesses are not short of data. If anything, they are drowning in it. The problem is fragmentation and interoperability – the ability of systems to connect, communicate and use information seamlessly with minimal human intervention.
Right now, a single guest probably exists across four or more systems, with table availability, sales and location data sitting elsewhere again.
No person could realistically synthesise all that information together.
But AI can – if MCP is enabled.
Where MCP unlocks real value
When AI has access to context, it starts impacting revenue rather than just efficiency.
Instead of generic marketing campaigns, AI can build highly specific segments based on real behaviour. Not just “visited in the last 30 days” but “high-spend guests who only visit weekends and haven’t returned in six weeks”.
From personalised communications to recognising returning guests, AI can enhance the human experience. It can let your team know about the guests coming into their venue today, what they might like, what to offer them or ideas for how to surprise and delight them.
To do any of these though, it must have access to the right context.
Where AI will not help
“Just ask Claude.” I’m sure everyone still reading this has heard or said that phrase recently. But even with MCP, AI is not a silver bullet.
Layer AI on top of disconnected data and it either does very little, or worse, makes decisions based on incomplete information and just makes stuff up. I’m sure everyone has experienced this after challenging ‘Claude’ on something recently.
Fundamentally, AI still lacks judgement. It can recommend and automate, but it cannot understand nuance in the way the best operators do. Hospitality is still a human business, and unchecked AI will lead to chaos.
Where to go from here
This should not be viewed purely as a technology conversation. It’s a strategic one.
The winners over the next three to five years will not simply be businesses that “use AI” but the ones building clean, structured and accessible data environments.
Pretty much everyone is in the same position right now. You do not need to “implement MCP” tomorrow, but you should start questioning where your data sits today, which systems talk to each other and most importantly speaking with your current and future tech providers around this.
There is a lot of noise around AI right now, but the reality is simple: AI will only ever be as good as the context it is given and ability to then speak directly to the systems being used in hospitality.
Model Context Protocol is one of the clearest signals yet of how that context will be delivered in the future.
For hospitality leaders, understanding MCP now could be the difference between experimenting with AI and actually extracting commercial value from it.
Carey Benn, CEO of Guestwise
