SOLUTION

Gathering data, insights & observations

First, I took peek at the current conversion. Between the two call-to-action buttons on the landing zone page, 9.2% of visitors were making it to the profile page. Not bad, but then only 2.2% of the profiling population was actually viewing the proposal page. I knew we could do better.

Second, I kicked-off discovery by revisiting the genesis of the current profiling screen. Three considerations heavily shaped the minimal viable product (MVP) version:

Then, I considered our primary (Trish) and secondary (Jennifer) personas:

Lastly, I reviewed existing customer research and verbatims specific to the profiling page:

Generative interviews

To validate user sentiment and generate ideas, I worked with our user research partners to run a one-day user workshop. We invited a dozen representatives from our target customer population to come share impressions of the onboarding flow. After the interviews wrapped up, we used an affinity diagram to chunk insights into themes. I then led the product team through an exercise to develop a clear problem statement to frame our exploration.

Generating ideas

As the design lead for this project, my approach was to set the stage for the entire design team to explore solutions. Once ready, I held exploration sessions with the design team to generate and refine ideas:

Narrowing down ideas

My next step was to create three testable prototypes representing the best solutions. Once the ideas were clearly expressed, I consulted with our development lead for input on the technical effort and considerations of each design. Then, I worked with our user research partners to formulate and run an unmoderated, remote test to evaluate each design.

Selecting the champion idea

The final step of the exploration phase was to present the three most promising solutions, technical considerations, and user test results to project leadership for discussion. The second option, the single question, stood out as the clear favorite among users and the product team and was soon dubbed the “Quick Path Profile.”

Appealing to legal

Typically, a product team would submit high-fidelity mock-ups of a proposed design to a generic legal inbox. House counsel would review the designs and reply with any objections or alterations in the form of a PDF. This back and forth was continued until both parties were satisfied (or one relented). We took a different approach:

Altogether, by having counsel fully present and working with the team, the project was able to produce a new product aimed at a new audience within the comfort zone of our legal partners.

Detailed design & build

Once we had a production candidate, the team conducted a lab study to validate the solution. One of the tricky design details we needed to sort out was how to indicate the user’s location and progress in the flow. For this test, we used a simple “Question X of Y” to display progress. Although this approach was straightforward, it was imprecise since the total number of questions can vary. For example, if a user indicated the account was for both them and a partner, we inserted an additional question that asked their partner’s age. A few, keen-eyed participants called this out when the total number of questions increased as they progressed in the flow.

As users interacted with the prototypes, I observed how swiftly they made their way through the flow. What if the ease and speed of completing the flow removed the need for a step indicator? Given that we were testing a coded prototype, it was simple enough to comment out the status indicator. As suspected, users completed the flow quickly and never hesitated to question where they were in the process. So, we dropped the status indicator for production.

We also introduced an additional way to access the Quick Path profile. Two "Get Started" buttons (one located in the hero image and the other at the bottom of the page), provided entry to profiling on the MVP landing zone. For this release, we replaced the bottom "Get Started" button with the first profiling question. The hypothesis is that an easy-to-answer, interactive question would entice users to fill in the box to see what would happen.