Messaging in the middle of the sea.
Client: Royal Caribbean
Role: Product Design Manager
Date: Dec 2018
Guest to Guest Chat
For years, people have been going on cruises in groups: family, friends, or even strangers who meet on the ship. But getting in touch in the middle of the ocean where there isn’t access to a phone or internet can be pretty complicated. This is why guest to guest chat is one of the most requested app features across the Royal Caribbean fleet.
Research & Ideation
Since our messaging feature would be integrated into the larger Royal Caribbean mobile app, we formulated our personas based on the data we gathered from guests on our ships. Understanding that our guests look for familiarity and ease of use onboard, we took a deep dive into the messaging apps already out there to outline key features. The biggest challenge was accounting for privacy, yet allowing guests to easily message each other. Armed with that knowledge, we did quite a bit of whiteboarding, ideation, and story mapping sessions to plan the specific features and flows we wanted to focus on.
Bringing it to Life
As manager to one designer on this project, I was pretty hands-on in both the design direction and getting into the pixels when necessary. I steered conversations, formulated the strategy in close partnership with product managers, and collaborated with our engineering and data partners daily.
Initially, we focused on getting the flows and features right. By stripping away all the unnecessary details, we were able to hone in on key flows and use cases. The most crucial bits to nail down were guest lookup, group chat, and notification strategy. All of these relied heavily on access to guest account data and ship WiFi connectivity. Once those were defined, we moved onto crafting an on-brand yet familiar visual design.
Introducing Conversational AI
Capitalizing on the conversation, we introduced a chatbot into to connect even deeper with our guests. Personified by the ship itself, the bot brought a friendly voice to our app's existing featured. While our data scientists taught the bot our guests' most asked questions, we crafted a voice and visual style that could integrate seamlessly into our app. Isolating specific tasks, we crafted a response matrix to give guests the most useful response, next best action, and a set of "answer before you ask" prompts to continue the conversation.
Prototypes & Testing
Throughout the course of the project, my design partner and I built lo-fi and hi-fi prototypes using Framer, Principle, and InVision. These allowed us to test remotely and with guests onboard at higher fidelity. Not only were they useful in communicating ideas to guests and stakeholders, but also to fill in the blanks for anything we might have missed while creating static comps.