Project background

Have you ever tried to pick up a rental car without a credit card? If you have, I feel your pain. If you haven’t, first imagine that you got a discount by pre-paying for a nonrefundable booking. Then when you show up to get your car, you're required to prove your identity and address as well as income, bank balance, or credit score. Requirements vary wildly so good luck knowing what documents to bring. If you don't, you'll get turned away and instead of going on your trip, you have to appeal for a refund and find other transportation.

The scenario represents a small percentage of transactions but more importantly, behind every one is a human being having a truly terrible time. 

As the car product team, we asked ourselves "How might we... prevent these disasters from happening?" We looked for opportunities upstream in the checkout process to help customers make more informed purchases.

Discovery

• We listened to customer service call recordings and heard real-life use cases and frustrations.
• I interviewed head of the cx call centers to reveal root causes of customer pain points and understand agent pain points.

• Then, I audited the checkout flow for usability and the content for clarity.

• As the product team, we had little influence on the pick up experience, but we concluded that better ux during check out could prevent mishaps downstream. If customers understood the pick up requirements before they purchased they could avoid booking a car they wouldn't be eligible for, or come to pick up prepared.

Hypothesis

If we deliberately add friction to slow down the check out flow and progressively reveal the restrictions, customers would self-select out or be better prepared for pick up. As a result, the denial-of-service rate will decrease and customer pain will decrease.

New UX Prototype

User test prototype

User testing methodology

• Usertesting.com
• Screen for participants who:
  • do not have credit cards
  • have rented a car in the past two years
  • live within 1 hour of a major airport
  • hold a US passport
  • are comfortable making purchases online
• 5 Participants to test "Credit-only" flow
• 5 participants to test "Debit with restrictions" flow
• Clickable Invision prototype for desktop

Video clip of test participant's final thoughts [new window] →

Test Results and analysis

• Credit-only flow
  • 4/5 participants correctly understood the restrictions
  • 4/5 understood all UI affordances (expand/collapse terms, CTA disabled until terms are accepted, etc.)
• Debit-with-restrictions flow
  • 5/5 could name pickup requirements
  • 4/5 understood pickup requirements correctly
  • 3/5 correctly understood non-refundable policy
  • Understanding of UI affordances varied (card holder name pre-populated and read-only, expand/collapse terms, CTA disabled until terms are accepted, etc.)

Conclusion and next steps

• Iterate with more visual emphasis on refund policy
• Debit-with-restrictions flow has very heavy cognitive load
• Pay particular attention to formatting information and providing visual cues
• Allocate adequate design time to ensure UI is comprehensible

Product Test strategy

After iterating on the prototype, we launched an MVP to get a quick read. We expected that transactions would drop somewhat because of ineligibles and added friction but hoped that they would be offset by fewer refunds.

KPIs

• 90 days: Denial-of-service cancellations should drop
• 90 days: Customer service calls by debit card customers should drop
• 120 days: NPS among debit card customers should increase

Final ux

Demo - Screen recording of CX manager demonstrating flow to customer service agents

results

Huge win for customers!

• 33% of the customers in the A/B test changed to credit card, 60% continued with debit card, 10% searched again with the ‘debit friendly’ filter

• Reduced denial-of-service cancellations by 14%

• Reduced volume of this type of customer service call by 55%