Wanderu search cards

Wanderu is a Boston based travel startup focused on empowering users to easily compare, book, and save on bus and train tickets throughout North America and Europe. Between the apps and the website, Wanderu has millions of active users.

Empathizing with the user

While running some usability tests on our standard search flow for Wanderu’s conversion team, we realized participants were missing key details about the trips they were selecting. They were getting confused about the mobile boarding icon. There was a tendency to miss the number of transfers. Participants who selected trips with transfers proceeded down the funnel only to realize that the trip selected was not an optimal route. Additionally, participants were frustrated that in order to find out more information about their trip they would have to proceed further down the funnel.

Defining the problem

Our hypothesis was by organizing the information in a more clear manner there would be less cognitive load, and by bringing some of the information further up the funnel into the search cards there would be less user friction from going back and forth between search results and trip summary. With both of these improvements we expected to see a substantial increase in conversion.

Ideation

We explored numerous ways to format the time, station and duration on the cards. We then explored various mechanisms to expand the card. These varied from having an expand icon, to having a button that said “x transfers” to a button that said “details”.

We did numerous explorations of type treatments for different attributes such as time, price, duration, station, transfers, ticket type, vehicle type and more. Typography was a focus, to instill clarity from an information hierarchy stand point. Through type we hoped that users would be able to easily scan and parse the necessary information.

Prototyping & testing

We built a few prototypes and tests to have participants compare a number of trips. By having them compare the trips we could observe things such as scannability, we could see if they expanded the card for more details, and we could see what information users prioritized. We designed our test to show participants a bunch of really bad trips with multiple transfers, or at bad times of the day, and one ok trip. We hypothesized that if the new card variations were better, more users would notice trips with transfers. Conducting multiple usability tests comparing the variants against the original, we noticed some correlations that seemed to validate our hypothesis. With the new expandable versions of the card, less users bounced forward and backwards in the funnel, and more participants were noticing details like if there was a transfer or not. Participants with the old version tended not to notice the number of transfers as much and were more likely to proceed to trip summary only to go back once they noticed the transfer. Here is the design we chose to move forward with.

These designs were split tested on the live site. 20% of traffic received the new variant and 20% of traffic was reserved as a control group. After testing on the live site for a week, the new variant was the clear winner. Our hypothesis seemed at least partially validated. We saw a ~10% improvement in overall conversion rate of the site. We also saw users spending less time on the trip summary page. Prior to implementing the expandable cards, we were showing the same information in a slightly different way causing unnecessary cognitive load on the users. With the expandable cards, users who wanted to see more information could expand and see, and users who had simple one way trips with no transfers (which is the majority of trips sold on wanderu) could move on quickly to checkout.