San Kim
thumbnail.png

Auto Equity Loan - Opt-in Page Redesign

 

LendingClub

 "How can we provide more clarity about the loan to help applicants decide if it’s the right choice for their situation?" 

 

LendingClub, now an online banking platform, was a peer-to-peer lending company, where people can borrow money through the platform and everyday investors can fund those loans and make interest in doing so.

On the borrower-facing platform, Auto Equity Loan was launched, where people can use a fully paid car to borrow money against. As a MVP product, it was only offered to people who didn’t qualify for a personal loan due to low credit score.

 

Challenge

The existing opt-in experience for the Auto Equity Loan had a low conversion rate of 20%, with a high number of unqualified applicants opting into the loan, resulting in a low offer take rate.

The team set out to run an opt-in page redesign experiment to maximize our learning, increase the conversion rate by 10%, and increase the number of qualified applicants submitting their car information by 5%.

Outcome

I designed three variants that we tested against the existing opt-in experience, with each variant informed by and designed to address specific user pain points identified in user testing.

All three variants increased the conversion rate by 7-10% and the vehicle submission rate by 1-6%, validating the improved user comprehension of the loan product and its details for making an informed decision. At the end of the experiment, we replaced the existing opt-in page with the best-performing design variant.

 
 
 

Team: 1 PM, 3 Engineers, 1 Designer | Role: End-to-end product designer | Platform: Mobile web application

 
 
 
 

Research & Findings

  • Unmoderated user testing on the current experience

    • 10 participants

    • Screener questions

    • Interactive prototype

    • Final questionnaire

To understand how applicants perceive the existing opt-in experience of the Auto Equity Loan, I used UserZoom to recruit 10 participants with similar attributes to our actual applicants, such as credit score, income, intent, and age.

After they went through the prototype of the existing opt-in experience, we asked questions to evaluate their level of understanding of the product and identify the main barriers to opting into the loan product.

 
 
 
 
 
 
 

Design Process

  • Design variants A/B/C

    • Variant A: Maximize loan comprehension

    • Variant B: Maximize comprehension of Personal Loan decline

    • Variant C: Seamless transition from Personal Loan

Once in the design phase of the project, I aligned the designs with the user needs discovered from the user testing and identified three main design goals to ground my design variants in:

Design variant A: Help applicants understand what Auto Equity Loan is so that they know what is being offered to them so they feel comfortable opting-in to learn more about the loan.

Design variant B: Maximize the applicant’s understanding that they’ve been declined for a personal loan, emphasizing the urgency of considering the alternative Auto Equity Loan

Design variant C: Make a seamless transition from the personal loan application to the Auto Equity Loan application to keep the focus on helping applicants get the financial assistance they need.

 
 

Design variant iterations

 
 

Experiment Results

 

Once we finalized the design variants A, B, and C, we began the experiment by funneling 25% of all incoming applicants into each variant to observe their performance against each other in terms of conversion rate and vehicle submission rate, which is the first step of the Auto Equity Loan application process.

Two weeks into the experiment, we observed that variants A and C outperformed the existing opt-in page in conversion rate. With the success of variants A and C, we adjusted the traffic distribution to funnel 5% into the control design and split the remaining traffic between variants A and C. At the end of the experiment, we replaced the existing opt-in page completely with the best-performing variant, C.

 
 

Experiment results