Iterations
Low to high-fidelity Prototype & User Testing
We designed mid-fidelity prototypes to maximise the clarity of user flow while minimising time and effort. I was in charge of the analysis and profile page while helping to modify the onboarding and trigger-logging flow. After presenting our prototype to 8 participants (one conducted by me), we uncovered several usability issues and iterated upon areas of confusion, and I converted the insights into the final high-fidelity prototype.
Homepage
More than half of the participants mentioned that they were unclear about the function of the top-right button and the hide button. Half the participants also expressed confusion interpreting the forecast graph. In response, we decided to use intuitive iconography and language, remove the hide personal detail button, and improve the clarity of the graph using legend and colours.
Symptom logging
When seeing the body diagram, most users immediately questioned whether it was the front or back of the body. They also requested more flexibility and precision for marking flare-ups, such as around the joints and creases. Thus, we changed the marking mechanism to painting instead of selecting and replaced the 2D diagram with 3D visualisations.
A/B Testing for Trigger Logging
We tried to incorporate playfulness into the trigger-logging process by designing a card-swiping interaction. Not sure about the user acceptance, we conducted A/B testing with our users comparing the two methods. Users reported that multiple-choice logging is faster and more functional while gamified logging is more engaging. Although it was a tie between both methods, we eventually chose method A because one participant reported strong negative feelings toward method B claiming that ‘my condition is not a game’.
For our users to better understand the mechanism of the smart trigger recommendation, we designed an alert component to be displayed with the trigger log. However, almost all of our participants ignored this alert so we increase its visibility by prioritised placement and elevated colour. Users also want to be able to manually add a food or event.
Analysis page
We found that participants didn’t feel drawn to the mood-tracking feature so we removed it in the new version. Users were generally more interested in the potential triggers than known triggers, so we moved the list up to a more prominent position.