How we upped user confidence and comprehension four-fold,
and actually managed to make validation test users excited about a fake Egypt trip.
"When you showed the prototype, we looked at each other, and I could barely contain my smile. I guess that's the design process, the first time we saw it it was OK, then the next time it got better, but the final time, wow, we were so surprised."
- Zaky Prabowo
WeTravel Co-Founder & CMO
"Chia is incredibly driven and does not compromise quality. She is incredibly easy to work with, because she can work with great autonomy but will still consult management at the critical junctions of a project. It was a true pleasure to have her lead our project team and I can not recommend her highly enough."
- Johannes Koppel
WeTravel Co-Founder & CEO
WeTravel had a trip details page that didn't work. Users were getting confused by the booking process.
Here's the quick TL;DR of what we did.
WeTravel is an early stage startup with a mission to make it easier for people to plan and join group trips. Trip organizers use WeTravel to create itineraries, manage trip communication, collect money, and promote their trips. Travelers use the site to discover and join trips that organizers have posted.
As the leading designer, I managed a team of six designers and UX researchers. In addition to leading all aspects of the research, analysis, and prototyping, I also served as the point of contact for the clients. I was responsible for keeping the project on schedule, on budget, and aligned with the clients' business goals while advocating for the users when need be.
As a collective, we brainstormed the things most important to travelers and created questions for them.
Answering these questions will help us understand the users' emotions as they journey through the website.
The purpose of our Card Sort was to understand users’ existing mental models of information architecture and procedural flow when booking a group trip. We conducted 5 sessions of cardsorting, each session with groups of 2 or 3 users who have never used WeTravel previously.
Turns out there was just way too much information for users to juggle and sort coherently.
We built a medium-fidelity prototype, filled with various iterations of the pages and flow, and had some sample users try them out. (While we observed creepily and took notes).
Based on users' feedback and observed interactions with the WeTravel site, we identified numerous pain points, aspects of the site that were hindering task completion and caused confusion and frustration.
Turns out, usability testing was much more effective than cardsorting in this case.
We grouped pain points (each written on a Post-It) into buckets based on commonality. Through this, we discovered that our users experienced three main themes of challenges:
"I’m not sure if it’s actually happening… it’s vague."
"Will it immediately book? Or do I get to review the booking first?"
Modeled after Google design sprints, our design sprints aimed to give stakeholders context on user pain points and refresh them for team members while ideating on possible solutions.
We began the sprint with lightning talks - cliff note summaries of pain points and personas as fast as lightning.
Then, we did "Crazy 8's" - successive string of eight one-minute ideation sprints where we sketched all possible solutions to user pain points we could think of.
We even squeezed in a 15-second dance party.
With each new batch of mock-ups, we tested a new group of users. Four rounds of testing in four weeks, go-team!
Users could find all relevant trip information within 10 seconds of looking at the landing page.
4/5 users thought the site was trustworthy, compared to 0/5 in intial testing
"What you’re selling is meeting new people and going on a trip together."
"I see everything I need to feel confident to confirm [booking]."
This project was a delight. Special thanks to the founders for the amazing opportunity and collaboration.
Though that's it for this sprint, I'm sure there will be more redesigns in the future.
Design is neverending, afterall.