Three studies that reshaped a new member experience

The problem

Thrivent’s PLE team was building the post-login experience (PLE): everything a member sees after signing in, from their insurance and investments to documents and member programs.

The digital team was split into pre-login (the marketing pages people see before they sign in) and post-login, and post-login was split again between the website and the mobile app. Two different agencies had built them, on separate login systems that never shared a single identity for members, so the experiences had drifted apart. Before building, the team needed to know which page designs worked, whether people could find things in the new navigation, and what to call features.

My role and the team

I joined Thrivent as the third member of a small usability team. As the team’s senior researcher, I took the post-login experience, so its research requests came to me. I was the lead UX researcher on the mutual funds prototype test and the navigation tree test, and one of three researchers on the naming test. All three were run by the Usability Team I was part of, using the research plan and readout format I’d set up.

The approach

Each question got the method that fit it, all within a few months for the same team.

Two mutual fund designs, tested side by side. The team had two designs for a member’s mutual fund account: one scrolling page, and a tabbed layout. I wrote the research plan, with the assumptions to test, a screener, tasks and a script, then ran moderated sessions with 12 people who manage a mutual fund online and weren’t Thrivent members. Each person was interviewed about how they use their account, then tried both prototypes, in alternating order to avoid bias, on tasks like redeeming a fund and checking growth over time.

A tree test of the navigation. To see whether the proposed menu worked before it was built, I ran a tree test with 50 people who weren’t Thrivent members, on ten common tasks such as making a payment, finding statements and contacting a financial advisor.

A naming test. The team needed one term for a digital-only document experience. We asked people which term they preferred, then tested “Paperless” and “eDelivery” in a prototype, each with a separate group, measuring whether people completed the task directly.

What we learned

The labels didn’t match how people think about the products, and neither design won outright.

• Neither mutual fund design won. Preference split roughly evenly, because each design’s strengths were the other’s weaknesses. The scrolling page had a full-width chart but hid what was below the fold. The tabs made content easy to find but forced people off the page.
• “Protect” didn’t say insurance to participants, and some didn’t want to be pushed to contact a person when they could do it themselves.
• Nobody looked for their advisor under “Support”, so it didn’t belong there. FAQ was the hardest item to find, and charity and donations had no clear home in the menu.
• “Paperless” won on both counts: people preferred it, and it did better in the task test.

What changed

For mutual funds, I recommended a hybrid: keep one scrolling page and the full-width chart, and add navigation that jumps to each section. The team redesigned the page combining the best of both.

The tree test sent the team back to rework how the navigation was organized, and “Paperless” became the term for the digital-only experience. Not every recommendation from these studies shipped, but many did.

Looking back

When I joined, the team was shipping fast with almost no research behind it, so we started with quick evaluative studies of what designers were putting out. Those studies kept surfacing questions that belonged earlier, in strategy, which made sense for a team moving from agency work to building in-house.

I’d be bolder sooner about taking that evidence above the product managers, who were under pressure to deliver, to the directors setting strategy. When a director later saw one of my studies and asked me to present it at a department-wide all-hands, it showed how much those relationships widen what research gets to influence. I’d build them from the start.