Just keep testing, just keep testing

Stephanie Wang
MHCI Capstone 2020- Highmark
7 min readJul 20, 2020

Last sprint, we ended with a medium fidelity prototype. This sprint (our final one!), we tested that prototype, iterated upon it, and tested again. In both rounds of testing, we designed tasks and questions aimed at feature validation and usability.

What to do with onboarding: goodbye goals, hello flowers and pigs

When we originally asked members to set goals during onboarding, we wanted to prime them with their own existing internal motivations to stay healthy and save money, so that when they used our application, they would associate preventive tests with staying healthy and saving money.

When asked about the value of the goals, some participants thought they served as reminders, but others were apathetic to them.

“ Having the goals recorded in profile is a simple reminder; just to see it written out or said out loud makes the goal more real as opposed to a thought in my head.” — user testing participant

Although it is nice to have the goals as reminders, we wanted to better leverage onboarding to communicate the value of going to the doctors for preventive checkups for both staying healthy and saving money. Our current design does not make that statement. Furthermore, our client did not see business value in tracking personal goals, and even if there was use in tracking, it would’ve been difficult to track goals such as taking more regular walks without adding variables outside our current scope (e.g. including wearables).

We ideated on potential ways to communicate the idea that going to the doctors for preventive check-ups helps with both staying healthy and saving money. Our concepts included:

  • Changing the goals so they are trackable within our product and system (e.g. if their goal was to be more active, our app should be showing them their activity progress)
  • Using stories and facts to visualize the consequences of not going to the doctors
  • Visualizing cognitive dissonance through interactivity

We ultimately decided on going the interactive route.

We tested this concept in low-fidelity by using survey style questions, the testees understood the intent clearly.

“ You’re trying to get people to see that overall health and how much money you spend are linked. You’re also trying to get people to work on their preventive medicine more so than corrective.” — user testing participant

Thus, we designed an interactive playful version of onboarding in which people choose if a topic helps them stay healthy (flower) or save money (pig) or both or neither. Then we hone in on the punchline that getting preventive tests does both.

How “Book Test” button K.O’d “Remind Me” button

Originally, we included a “Remind Me” button in addition to the “Book Test” button in order to make the journey to book test a little easier. In case people didn’t want to book immediately, they had the option to remind themselves and come back to it later. However, when asked about the functionality of the buttons, in both rounds, most of our participants did not see much use in “Remind Me”.

“I’m fairly good at remembering when appointments are and I use a separate calendar to set reminders. You get a reminder from booking it anyways.” — user testing participant

What we decided was instead of having the member manually remind themselves, the system will automatically send members reminders to re-engage with the app.

Seeing family information is here to stay

One of the biggest features we were debating on is if we should include the aspect of seeing others’ tests. Originally our intent was that this would serve as nudge for others. In our previous user research, we found a trend that people would want to share the information with their parents and spouses.

In the first iteration, people got confused about where to find this. We quickly realized that calling this section “All Test” did not make much sense. When asked if they knew anyone that would find this useful, there was mixed feedback, with someone replying “honestly, no”.

Along with relabeling, in the new iteration we really focused this for a member’s family. Instead of filtering for anyone, this now only allows the addition of family members which narrows the scope of the feature and makes its purpose clearer to users. We got more positive feedback on usage after the redesign.

“I will gladly add my father, I will use this to give him the care he needs” — user testing participant

When asked if they would use it for anyone, people mentioned parents and spouses as we originally expected.

Nav nav nav: how to get around

In our first prototype, we had broken down the information into four main tabs. We observed that our first round testees had some challenges navigating, and they thought there was too much back and forth between tabs. In their minds, there was a lack of cohesiveness between the features of our design.

We ideated on ways to organize the information including:

  • Keeping all tabs except “All Tests” and embed that elsewhere
  • Using the hub and spoke way of navigating instead of the nav bar
  • Reducing to two main navigations on the bar

We ultimately decided on keeping the navigation bar but only having a dashboard and “My Tests”. As “My Tests” is the core functionality of our product, our intent with this choice is to keep that at the top of the hierarchy.

Groups groups groups: how to organize test hierarchy

One of the biggest challenges and discussions in terms of information architecture was how to organize the tests. At first, we focused on presenting the tests listed and were struggling with how to group them in an easy-to-use way. We originally attempted to organize the location of tests by filter tabs. However, when tested, participants had a lot of confusion with them.

We then tried not grouping the tests, which usability testing-wise was fine. But we wanted people to be able to get multiple tests at once so we realized that we should still prioritize where they can get the tests done.

One of the cases that was throwing us off was what would happen if a test could be done at multiple places. We finally resolved that by recommending one best location for the member. This also would take the burden off of the member.

Graphs graphs graphs: what is digestible and impactful

One of the main points we wanted to convey was the lifelong savings aspect of going to the doctors consistently for preventive care. Because of that, we have a graph showing lifetime monetary costs. However, in our initial testing, people were more confused than convinced.

“Am I saving dollars? How am I saving dollars? Is it because I’m living longer? Is it because my insurance is covering it?” — user testing participant

We redesigned the graph and gave it more context. Then we A/B tested two versions.

Most of the people we user-tested preferred Graph A because “it gives a clear breakdown as to how much I’d spend if I started testing and at what ages. It makes a bigger impact to me spending 50k over potentially over 45 more years is a lot less per year.

Another section with various graphs is biometrics. Our purpose was to show people aggregate data so that if they haven’t gotten the tests, they’d be curious about their own metrics (e.g. blood pressure and cholesterol level) to get the test, and if they have gotten the tests, they’d want to improve or maintain their metrics.

In the earlier version, we separated comparison and personal data. However, we realized that we could combine them to make the information more impactful and digestible since they are so related.

How “The Killers” escaped: team virtual game night

In the middle of the chaos, our team bonded through some virtual game night where we played Jackbox, skribb.io, and Codenames.

We learned that no one on our team can keep a straight face. We all went off-camera as spymasters. That’s one nice unintentional benefit of Zoom. We also discovered Corey’s talent at clue giving!

The end is in view

After all of our rounds of testing, we’re at the point of creating our high fidelity prototype. We came up with our design system and we’re getting yet another prototype ready for testing. The end is almost here though, stay tuned for our grand finale!

Stay safe,

The A-Team

--

--