User Research for a Mobile App Measuring Everyday Function

The Ambiguity

Subtle declines in everyday functional ability — the instrumental activities of daily living (IADLs) like managing medications, finances, or a schedule — can be an early warning sign of dementia, often showing up well in advance of a clinical diagnosis. However, there was no widely available tool for tracking objective IADL function.

Building and validating a tool like this is genuinely difficult, requiring a significant investment of time and effort resources. That’s why understanding whether older adults would accept and adopt the technology, and how they’d use it, was critical at the MVP stage. Getting that wrong early would have meant validating a tool that few or none would use.

Approach

The research moved in two connected phases.

Phase 1 was a beta-test early in the MVP stage. Twenty-four community-dwelling older adults (average age 71.6) were randomly assigned simulated IADL tasks to complete through the app at specified times, every day, for two weeks. The goal was to confirm basic functionality and surface what needed to change before recruiting users for the larger sample. We collected adherence data plus a survey and individual interview for each user at the end of the two weeks to inform that refinement.

Phase 2 — a larger sample, 131 community-dwelling midlife and older adults (average age 67), completed daily tasks through the refined app, then shared feedback through semi-structured interviews, with a thematic analysis of what shaped their comfort sharing performance data and their sense of the app’s value in a healthcare context.

As Principal Investigator on the parent study and a member of the app design team, I was involved on both the research and product sides. The interview and survey findings came back where design decisions were being made for future iterations.

What We Found

Phase 1 confirmed the app’s basic functionality held up in practice. Adherence was high, with an average total task completion rate of 94.2% across the two-week period, and follow-up interviews described the app as generally easy to use, with tasks that could be completed quickly. That result, plus the specific friction points identified in those same interviews, is what determined minor revisions to the app prior to transitioning into the larger Phase 2 study.

Phase 2’s thematic analysis then went deeper into the adoption question of trust. It revealed that people’s interest and confidence in using this kind of technology tracked with three specific beliefs: 1) the perceived value of having real-time information about their own cognitive performance, 2) their prior experiences with healthcare providers, and 3) their trust in the technology’s security and accuracy. General usability got people to complete the tasks; trust is what ultimately would determine whether they’d adopt a tool like this in real life and/or share the information with their healthcare providers.

What Changed

Phase 1 functioned as an MVP-stage beta test, confirming the app worked at a basic level and directly informing the refinements for the version tested in Phase 2. Being on the app creation team as well as the research team meant those Phase 1 findings fed straight back into design decisions, not through a handoff to a separate product team.

Phase 2’s trust-related findings now point to the next set of concrete design decisions — how performance data and its real-time value are communicated back to the user, how the tool integrates with or explains its relationship to a person’s healthcare provider, and how transparently it addresses data security and accuracy.