Rental Cars

Overview

Rental Cars on Tripadvisor launched in early 2019 and I had the privilege of taking over as lead designer in late 2019. Below is a collection of the many efforts we made to improve and optimize the product.

Rental Cars exceed growth projections by 37% in Q1 of 2020.

First, Some Context

Tripadvisor Rental Cars' interface was originally an iframe containing RentalCars.com's search experience. The team recreated this interface using Tripadvisor's design system to establish a baseline. After reaching parity with the iFrame experience across a number of metrics including booking per search, the in-house treatment was rolled out to 100% of site traffic.

Map View

Prior to me joining, the team had put together an MVP map view for search results. User tests showed travelers either struggling to browse results or ignoring the map entirely. We made a series of improvements to reduce pain-points and make the interface more assistive.

To prevent overwhelming the user with pins and and sharing below average deals, we limited the number of pins displayed. We surfaced only the top deals and displayed the remaining rentals as collapsed dots.

To allow users to view deal details without having to switch to the list, we introduced a popover state. We also condensed all deals within a location to a single pin, instead of having a pin for each deal.

In order to faciliate seamless transition between the map and list view, clicking on a pin highlights the deal in the list view and vice versa.

Finally, analytics showed that users who engaged in filters were more likely to book a rental. To streamline the process we added a horizontal set of filters above the results. Previously, the user had to click a button revealing a slide-in list of filters.

Assistive Calendar

The team was curious to see if travelers would find utility in a calendar that highlighted dates with the cheapest rates. More specifically, we wanted to know if users would adjust their travel dates to capitalize on car rental savings.

To test the concept, I put together a plan for an open-ended user survey. Participants were tasked with finding a car for an upcoming trip, and then asked how much they would have to save before they'd consider adjusting their travel dates. These were the takeaways.

• On average, users were very unlikely to adjust their travel dates, many requiring savings of $15-$500 dollars before considering it.
• Airfare and inflexible dates were cited as the main reasons for not adjusting their car rental booking dates.
• If users did adjust their car rental dates, it was typically because of a change in overall plans.
• Many users were willing to add $6+ to their daily rate if it meant more room or flexibility.
• Many users felt that they could eventually find a car that would suit their needs and budget if they looked long enough.

We ultimately recognized that users most likely wouldn't adjust their travel dates without offering significant savings.

Pick-Up Location List

Tripadvisor’s “Write a Review” flow deprioritized pick-up locations, serving up irrelevant results and making actual supplier locations hard to find. At a stage where collecting reviews was critical for the fledgling product, we needed somewhere to send users to find their rental car pick-up location and leave feedback.
 
We added the ability to search for pick-up locations within the product. As an added bonus, we were able to deeplink users to the search page if they came in via a marketing e-mail.

Lander

The majority of our traffic comes from search engines taking users to our landers, which are scoped to the location in the users search query. So for example, if a user searches for car rentals in London, all the content on the lander will be London-specific.

Active Deals Shelf

In the saturated market of rental cars, users are searching across multiple companies looking for the best deals. We came up with a visual treatment to highlight active deals in order to communicate that they’ll always find competitive rates on Tripadvisor.

Local Transportation

For prospective travelers, a car may not be necessary depending on the destination. As a way of generating trust, we listed every mode of transport within a destination. Users can decide to book with confidence knowing what each city offers.

Passenger Count

We’re currently testing to see if asking travelers for their passenger count upfront gets them relevant results more quickly, increasing their booking rate.

Project Learnings

Prepare to be surprised

I can’t count the number of times we ran a user or A/B test and were shocked by the results. I quickly recognized that I needed to check all my assumptions at the doors in order to build a great product.

It's a team effort

This group's success hinged on everyone in product, development, design, marketing, SEO, SEM, analytics, and many other groups coming together to support each other's efforts. None of this could have been achieved alone.

Back to top