SnapTravel is a travel booking company that offers competitive hotel deals by using a chatbot to book hotels. The company knew there were issues with the bot experience because many customers would go straight to the In-App Website and ignore the ability to chat with the bot to find hotel results. Our goal was the uncover areas of improvement and design solutions for this segment of the user journey.

We used a number of research methods to identify potential areas of improvement.

Competitive Analysis: Examined and interacted with similar bots in the travel booking space.

Bot Design Best Practices: Studied articles and bot design handbooks to create our own design guideline document.

Usability Testing Videos: Conducted user-testing sessions where users recorded themselves accomplishing specific tasks with the chatbot.

Agent Chats: Investigated the chat log records for when the bot had failed to serve a user’s needs and a human agent needed to step in.

We collected and discussed our findings, allowing us to identify clear and persistent issues. We discussed these findings with the SnapTravel design to prioritize the issues that were most important to solve. We used this information to instruct what areas we should focus our design efforts on.

On-Boarding Experience: Issues involving the tone of the bot and with users understanding what the bot was capable of.

Hotel Search and Filtering: Issues navigating the bot’s workflow, specifically around its primary functions of searching for hotels and filtering the results.

This was an iterative process where we would create some mock-ups and prototypes in Sketch and then go out and put them in front of people to test our designs. Overall, we conducted 9 user interviews to inform our designs.

These are some of the changes we made to address the issues we found.