Apaleo says MCP Server use rose 70% in three months, and names the hotels doing the work

Apaleo published figures on 28 October showing use of its MCP Server up 70% between July and September 2026, split as 23% from July to August and 38% from August to September. The number worth more than the percentage is the list of named customers and what they are actually doing with it, because that is the part most AI announcements leave out.

Apaleo launched what it says was the first MCP Server in the hotel PMS category in September 2025. MCP, the Model Context Protocol, gives an AI agent one consistent way to read and act on live property data: reservations, rates, availability, guest profiles. The agent connects once, works with the permissions of the person it acts for, and leaves an audit trail. Apaleo powers more than 2,200 properties in over 30 countries.

What the named hotels are running

Appenzeller Hotels, three properties in Switzerland, runs a daily revenue workflow in Claude that finds bookings it can upgrade so a lower room category frees up for resale. The group puts the saving at around two hours a week, on top of booking reconciliation and reporting.

Brera Serviced Apartments added its own MCP alongside Apaleo's and built a data lake combining revenue and cost data. Its IT team produced a forecast for both halves of the year, work that had previously taken three people two weeks.

Smartments amended about 2,000 reservations in a single action, which Apaleo estimates would otherwise have taken roughly 50 hours.

LikeStay uses the MCP Server to turn price lists into rate plans, analyse reservation patterns and shorten PMS migrations, moving data onto its Apaleo stack in a few hours. McDreams Hotels, a German mid-market group, uses agents for repetitive processes and for daily data questions that used to need a report.

"We can automate repetitive processes, get relevant information faster," said Christoph Klein, chief marketing officer at McDreams Hotels.

"Once an operator can point an agent at live reservations, rates and guest data, the work gets done," said Apaleo founder Philip von Ditfurth.

The context Apaleo includes, and what it means

To its credit, Apaleo cites research that cuts against the enthusiasm. The h2c Hotel Chain Tech Report 2026 found 91% of hotel chains now use AI, up from 78% a year earlier, but only 15% have a standardised AI data access layer such as MCP, and 52% have no plans to adopt one. Integration with existing systems is a barrier for 38%. Reliance on AI rose from 4.7 to 5.4 out of 10 over the year while trust stayed flat at 6.5. External tools such as ChatGPT account for 60% of day to day AI use, against 27% for AI built into vendor systems.

That last split is the real argument here. If most AI work is already happening in chat tools outside the PMS, the question for any vendor is whether staff can point those tools at live property data safely, or whether they end up pasting exports into a chat window. Apaleo's pitch is that operators choose and swap their own model rather than taking the assistant the vendor picked.

What to check before reading across

The percentages are Apaleo's own and come without absolute numbers, so a 70% rise could sit on a small base. The case studies are self-reported and not independently verified. And all of them are staff-facing: revenue work, reconciliation, forecasting, bulk edits. None is a guest-facing agent, which is the harder problem and the one where a wrong answer reaches a guest rather than a report.

Apaleo does not publish pricing or plan requirements for the MCP Server, and gives no setup steps beyond pointing hospitality businesses at its MCP group in the Apaleo Community. If you run on Apaleo and want to try the Appenzeller workflow, that is the starting point, and the permission model is the thing to read first: an agent acting with a user's rights can do whatever that user can do.