Recommendations Game

A new hotel concept wanted to center its brand on modern expressions of omotenashi, the Japanese concept of anticipatory hospitality. A crucial aspect of omotenashi is relationship building – staff can anticipate your individual needs because they know you so well. So, to encourage connections and further personalize the hotel’s local recommendations, staff would be invited to bring their full personalities to work. Are you curious about seeing local music while you’re here? Talk to Tetsuya, the bartender. He’s plugged into that scene. Want to spend your time shopping? See Sayuri in guest services. She grew up working in her mother’s boutique and knows some hidden gems all over town.

Both management and staff were eager to pursue this concept but were skeptical about how it would scale. We pitched the idea of a digital tool that would allow them to craft itineraries for guests and, over time, build a library of personal, local recommendations. If a guest loved Tetsuya’s nightclub suggestion, concierge would record that at checkout, and the system would learn guest preferences over time. This sounded cool, but the client couldn’t picture the mechanics. They weren’t familiar with how recommendation engines work, so all they took from our pitch was that they should get an AI to replace their staff.

The outdoor 'yatai' restaurants of Fukuoka, Japan.

To explain the vision, a front-end developer and I teamed up to build an interactive recommender game. Each round, players would greet a simulated customer, ask a few questions, and then make a recommendation by selecting from the hotel’s existing library of local experiences. Over three stages, we showed how the recommendation engine learned patterns with new interactions and that rather than replacing their staff, the algorithm’s performance was stronger with them in control.

The game worked. The staff loved playing it; afterward, they were all clamoring to add more recommendations to the library. Everyone wanted their haunts in the system. Management saw how their vision was achievable and chose to invest in developing the tool. The IDEO team and I also came away with a stellar list of ramen shops.