Removing anxiety from lending

What shipped

NetPayAdvance offers online personal loans across the US. The application flow is where most users decide whether they trust the company enough to continue.

It sounds simple — fill out a form, get a loan. It wasn't. The existing flow had high drop-off rates, a legacy UI that felt untrustworthy, and a backend that sometimes failed to send verification codes. It took many iterations and rounds of copy updates before the redesign felt right.

Three problems to solve

I set three targets before starting:

• Conversion — positive growth in completions and fewer drop-offs
• Credibility — modernize the legacy UI so the product feels trustworthy
• Efficiency — users complete the application faster with less friction

Start with account creation

I kept the phone number field at the top because most users already expected it there. Added a progress bar so people could see how many steps were left. Below the form on the two-factor authentication screen, I placed user reviews — because that's where we ask for the most sensitive information, and social proof matters most at the moment of highest hesitation.

Authentication was the choke point

Mouseflow showed this was where users struggled most. After reviewing session recordings, the issues came down to two things: the system sometimes failed to send verification codes (a backend problem I couldn't solve with design), and users dropped off when asked for personal information.

I couldn't fix the backend. But I could reduce the friction on my side.

Raised trust at the highest-drop step

I added a trust badge and user testimonials at the bottom of the authentication step. When someone is handing their phone number to a lending site, a form field isn't enough. They need to see that other people did this and it was fine.

Personal info without panic

This page asked for essential information including a social security number. Users were leaving — not because the page was confusing, but because it felt risky.

I added a tooltip modal explaining why each piece of sensitive information was needed and how it's protected. The tooltip didn't reduce the number of fields. But it reduced the anxiety around filling them out, which was the actual reason people were leaving.

Income logic got messy

This section had three different paths depending on what the user selected. Multiple buttons affected which questions appeared next, which created a prototyping challenge — a lot of Figma frames to handle every combination.

I solved it by using components for selectable dropdown items. That cut the frame count, made the prototype maintainable, and made the interactive version feel much closer to the real product.

Keep stakeholders on the real decision

While reviewing four flow options, a stakeholder got fixated on the design of a specific component instead of the flow order — which was the actual decision we were making. I had to redirect the conversation back to the right level of abstraction without dismissing their concern.

That's a skill nobody teaches you. Knowing when a stakeholder is solving the wrong problem, and being able to gently refocus them, is as important as the design itself.

Results

The redesigned flow shipped to production and replaced the legacy experience. Post-launch Mouseflow data showed fewer drop-offs at the authentication and personal info steps — the two highest-friction points I targeted.

The trust badge and testimonials at authentication had the clearest impact. Users spent less time hesitating on that screen, and support tickets related to account creation decreased.

What I learned

Two things.

First, prototyping less is more. I wanted to build a 1:1 replica of the application in Figma — something that felt exactly like using the real thing. But that confused stakeholders who weren't familiar with the tool. I learned to show only the essential steps and let the prototype communicate intent, not simulate the entire product.

Second, trust is a design material. In lending, users aren't just evaluating whether the UI is usable — they're evaluating whether the company is safe. Every design decision is a trust decision: where the social proof goes, how you explain why you need a social security number, whether the progress bar tells users what's ahead. Usability and trust aren't separate problems. They're the same problem.