Case study with a strategic decision · ~90 min
Deploying an AI Guest-Service Assistant: The Meridian Hotels Decision
Context
Meridian Hotels, a mid-scale European group of 40 properties, has spent six months piloting an AI guest-service assistant — a chat-and-voice system that handles bookings, requests, upsells and concierge questions 24/7 in 12 languages. The pilot results are in, and the board wants a recommendation at next week's strategy review: scale it group-wide, keep it limited, or stop it. You are advising the Chief Operating Officer. You have the pilot data below.
The pilot data
| Cost per handled request | −38% vs. human front desk |
| Guest satisfaction — under-35 guests | +11 points |
| Guest satisfaction — over-55 guests | −9 points (“I wanted a person”) |
| Front-desk hours affected | ~30% of routine tasks automatable |
| Data & privacy | Stores conversation + booking history; consent flow still manual |
| Energy footprint | Cloud inference adds ~6% to digital energy use |