They often found it hard to find what they needed. The application's information architecture was fraught with navigational blackholes that were hard to recover from. And TripAdvisor's hotel, attractions, and restaurant bookings were of the utmost importance to the business's bottom line.
We needed to provide a better platform for customers to find what they need and what they didn't know they needed through machine learning and data science. This was a nearly year-long reimagining of home, search, and discovery.
I collaborated with another designer and three product managers throughout the year. We conducted user interviews and usability tests with customers, built prototypes, and iterated on designs every week. To help drive the initial direction, we piggy-backed off a series of tests we had done with personalization of results.
We coordinated with product and engineering to lay out a roadmap, build it in phases, and roll out to limited segments of our audience to monitor performance. Through this phased approach, we were able to learn through following the data and listening to our customers.
The information and results displayed changed with user input (e.g. "I'm traveling to Los Angeles on these dates") or location detection (e.g. we know the customer is at home in Oakland, so show this relevant content). We also used this as opportunity to show sponsored and recommended content through machine learning.
I built a simple prototype in Principle to design some interaction and animation ideas, as well as do some light usability testing during one of the iterations. The final delivered product was altered slightly from this.
We saw a lift of ~2.5% in hotel bookings, ~4% in restaurant reservations, and ~3.5% in attractions bookings. Weekly active users and average session lengths also increased significantly from the previous generation of the home page.