Bena’s Design Portfolio

Context

Problem
Credible is a rate comparison site that helps customers find and apply for the best loan option from multiple lenders. Our big design challenge was figure out how to get people to trust us with their personal information so we could provide them with prequalified rates.

Impact of this problem
Users have shown intent by clicking a CTA that leads them to the application, but drop off immediately or in the middle of the application.

How do we judge success

Run A/B tests on the baseline design and new proposed design to see if more users are completing and submitting the form.

Possible design solutions

Current Implementation: All 18 form fields are shown on one page.

Proposed Form A. Single field per page.
Ask only one question per page. This will allow for more context and conversational language.

Proposed Form B. Divide form into 3 sections.
Break the form into three pages with less sensitive questions on the first page.

A. A/B tests for the form

What we already know

I conducted user research on the existing form (all form fields on one page) and discovered that there’s a segment of users who scroll all the way to the end of the form and abandon and users who start to fill out the form but drop off at a specific field such as phone number.

We also knew through analytics and feedback from our customer support team that the most frequently clicked tooltips were for the fields: phone number, household income, and loan balance.

What we need to answer

How do people feel when they are asked sensitive fields? Refinancing student loans can seem like a daunting process and I’m interested in discovering if users are more interested in efficiency--getting through the form as fast as possible or understanding the process, even if it takes longer. Are people abandoning because they've encountered a sensitive question or because they don’t have the information on hand and need to find it somewhere else?

Research process

Methodology
I used a within-subjects study design (users completed both application designs but were randomly assigned to begin with Design A or Design B) and asked to think out loud as they went through the usability tests. They also completed a post-interview and survey afterwards.

Participants
6 users (4 females and 2 males) between the ages of 18 and 27 participated in the study. Four were undergraduates and two were graduate students. All of them currently have student loans and are attending college in a large metropolitan city. They are tech savvy and frequently use the internet.

Limitations of study
The sample size is not representative of our user base. Participants in this study skew towards students aged 21-24 attending an urban public college on the west coast. However, the purpose of this study was to test the usability of the design in order to gauge which layout engineers would build first. Engineering would still build both designs to validate the findings but due to engineering constraints, we wanted to know if we should allocate resources to build Form A or B first.

Most of the prototyping tools didn't fit my needs so I created HTML versions of the page with form validation.

Results

UX copy can help build trust
One interesting takeaway was the power of language in building trust. For the 'Expected graduation/completion date' question, I added the comment “Sending you good vibes” after users added their date. It elicited positive feedback such as, “Aw, that's cute” and “This is nice.” For the 'Loan balance' field, I included “Let us know how much you'll need and we'll work our magic =)”. This also generated smiles and chuckles; however, one user commented, “I don't know if I like emoticons handling something about loans or money. Too informal.”

Finance can be intimidating for many people and injecting some human elements reduces the tension and can make users feel more comfortable. However, it has to come at the right context. When there’s a moment of confusion for the user, it can raise a red flag about the legitimacy of the site. This erosion of trust happened with the example of being too informal during a serious financial question.

Help text needs to be helpful
For finance-related questions such as Loan Balance, users wanted to know HOW to answer it correctly. For sensitive fields such as a phone number, users were more interested in WHY we needed that information.

Discovery of two types of user personas
4 people preferred the multi step page and 2 people preferred the single form per page. Since users completed both forms, they were able to make a direct comparison and everyone agreed that single field form felt longer.

Two camps of people emerged:
1. Those who wanted to get through the form quickly. These people were also less likely to read through the help text and more frustrated by the time it took to finish the other design.
2. Those who want to be walked through the process. “It feels more personalized. Step by step feels more like what would happen if you went to a loan officer.”

One interesting approach would be to allow users to choose between a guided or no frills approach when filling out the application.

Next steps
We decided to move forward with the 3-step form and incorporated some delightful copy. We only included it when we thought there might be user fatigue as reading additional copy would also slow down the experience. I finalized the designs and built high fidelity mocks for the engineering team.

We A/B tested the new design against the baseline design.The 3-step form increased desktop conversion by 10% and lifted mobile conversion by 40%. Huzzah!