Discovery

During this period, I wanted to gain insights into how people approached and used insurance account homepages and navigations.

Some target questions include:

• What role does the member site play in our member’s health journeys?

• Out of all the features on our site, which ones are the most important to members?

• What features belong on the home page compared to in the navigation?

Participants

An account home page is something that will be used by a wide range of people with varying familiarity and across lines of business. Because of this, throughout this entire process I decided to recruit participants from the general population, with an even spread of individuals with Commercial, Medicare, and Medicaid plans.

Methodology and Analysis

In order to get a reliable quantitative measure of importance and preference, I used created surveys through Userzoom. In addition, I used closed card sorts to gather feature priority data.

Result synthesis was done through Miro and Userzoom

Key Findings:

What do people log in for?

1. Claims/payments

2. Checking what’s covered

3. Check prescriptions, appointments, and test results

4. Find a doctor

Feature priority

Participants sorted a list of features available on our member site into 4 categories:

• Very important (16 features)

• Nice to have (8 features)

• Less important (6 features)

• Not important to me (2 features)

What belongs where

Gained a list of what participants viewed should go on a home page compared to within a navigational menu.

Competitive Analysis

In order to stay relevant in the health insurance market and see what UPMC might be missing, I conducted a competitive analysis with United, Highmark, and Anthem. Instead of just noting what each company did and following suit, I wanted to understand the successes and pain points of each competitor to give a confident recommendation of what approach to take with the designs. Again, I recruited general population participants with a range of insurance type through Userzoom to evaluate the competitor pages.

Some coworkers generously shared screenshots of their insurance account homepages from these competitors, which I used during this test.

Key Findings:

What worked well?

Visual display of deductible

This was well liked across all competitors, and consistently top rated as one of the most important things on the homepage. Graphical representations were especially well received!

What didn’t work well?

Duplicate or specialized naming

Participants expected the navigational menu to offer new features that were not already available on the homepage. They also became confused at specialized names for features such as “Smartshopper” and “Spotlight”.

Clearly differentiated navigation categories

Navigational categories need to be named in a way to clearly show that they are different from one another. For example, a competitor that had categories called “care” and “support” was negatively received.

After handing off these results, the designers got to work crafting a new home page!

Evaluating new designs

The designers came back with a couple new designs for the home page and navigation.

The two home page options were essentially the same format, but with different pieces of information in the top 3 column section. The “info” top version has more account related information, such as primary care physician details and wellness rewards. The “activity” top version has more spending related information, such as spending summary and claims.

There were also two different navigation options, a traditional megamenu and a more unique “split navigation style” where the secondary navigation is separated below a banner image.

During this round of research, I focused on testing perceptions and preferences around these new designs. My goal was to find out if this new approach improved the experience, and what needs refinement.

Although I would have loved to conduct member interviews, time restraints meant that I would have to rely instead on Userzoom and their general population panel. Instead, I utilized the recorded “talk-out-loud” feature of Userzoom to still be able to gather some in-depth qualitative data from usability testing.

Results were again synthesized through Miro and Userzoom

Key Findings:

1st Impressions

Overall, the new home page designs have much better first impressions compared to the current home page. Using a modified version of the Microsoft Desirability Toolkit, I evaluated how the current home page and new designs were described. To the right, you can see that there was a significant improvement in the types of adjectives used to describe the new designs.

Activity vs. Info Top

Between the two versions of the homepage the designers came up with, the “activity top” version was perceived as more comprehensive and personalized. However, when participants were asked to create their own homepage by placing different pieces of information in available spots, they created a homepage more similar to the “info top” version.

Cards worked well

Taking inspiration from what worked well on competitor sites, the new designs feature a card layout on the home page to separate different pieces of information. This approach was well received and contributed towards the increase in positive adjectives used to describe the new designs.

What people say is not what they need

When A/B testing the two navigation versions, most people said they preferred the “split” version, but usability recordings indicated that the megamenu performed much better.

After presenting these results, the designers made refinements where necessary and worked on creating a high fidelity homepage design.

Information Architecture

After finishing up research on the home page, I turned my focus to redesigning the information architecture of the entire member site. Because several new features will be added in the coming years and several other features are being removed, the IA of the site needed re-evaluation.

Open Card Sort

I first conducted an open card sort to determine overall groupings of features. Participants were given descriptions of all the features and asked to group them in whatever way they thought made the most sense. These results were synthesized using Userzoom’s built in card sort analysis tool, as well as creating a visual representation of feature connections in Miro.

Results:

Around 7-10 major groups were created. After discussing with the designers, we determined 7 groups to be the better route. Broadly speaking, we ended up with:

• Money/finance group

• Insurance plan group

• Care group

• Wellness group

• Account/settings group

• Inbox group

• Contact us group

Closed Card Sort

Next, I conducted a closed card sort to validate my findings from the open card sort and to evaluate names for groupings

The designers came up with preliminary names for each of the 7 groups, which were used during this test:

• Spending and Claims

• Coverage and Benefits

• Care

• Wellness

• Inbox

• Profile

Participants were given descriptions of all the features and asked to organize them into the 7 groups above.

Results:

Through the course of the testing, “Spending and Claims” was changed to “Spending and Billing” due to people initially placing “Forms” and “Documents” into this group when they weren’t related to anything else in the category.

There were also some features that shifted categories:

Document Archive

Coverage and Benefits → Inbox

Why?

• People think of email when they see “Inbox”

• Email is where they receive things

• Documents that they receive from their insurance company should be in “Inbox”

ID Cards

Coverage and Benefits → Profile

Why?

• People think ID cards are personal information

• Personal information should belong in “Profile”

At this point, we had an interim information architecture outline:

Feature Label Testing

Over the last few years, many core features have been added to the member site without proper thought into naming, and there are many planned features that are currently unnamed. Because of these changes, the designers wanted to re-evaluate the names of all features within the site.

To do this, I ran a baseline test to evaluate the effectiveness the current feature names had. This was done by presenting the feature name, and asking participants to describe what they believed the feature to be/do. I then ran two feature name creation tests in which I asked participants to create names based on descriptions of the features.

Results:

Ultimately, there were 14 features that needed no change in their name, 8 features that could potentially improve understanding with a name change, and 12 features that were recommended to have a name change. To finalize the names, I ran a workshop with the designers where I presented user generated names and we collaborated to decide on each new name.

Tree Testing

Finally, I ran a tree test to validate the IA and label testing of key features that were either high usage, had shifted categories during previous testing, or the team wanted to double check.

Results:

Most features performed well, but a follow-up survey was run for the features that were problematic. Coming out of follow-up testing, there were some further name adjustments. The largest change coming out of this was the decision to create a new category specifically for “Documents and Forms” to enhance findability.

Previous Homepage Vs. New Homepage

As of September 2024, the new homepage and navigation have been launched on our Member Site!

Previous Navigation Vs. New Navigation