top of page
From Surveys to Interviews:
Uncovering Care Platform Opportunities
Background
Health2Sync’s Patient Management Platform (PMP) is a clinic-facing platform that helps physicians, nurses, and dietitians track follow-ups and patient data.
The roadmap focused on new features: the product kept growing, but no one looked back at existing ones. So I initiated this research to find gaps between the current experience and how healthcare providers actually work, and turned them into a prioritized UX backlog the team could act on.
The research ran in two phases: a screening survey on roles, clinic sizes, and usage patterns (40 valid responses), followed by 12 in-depth interviews with lower-satisfaction respondents to focus on high-value opportunities.
Problem Exploration
Before the in-depth interviews, we ran a survey covering “usage frequency,” “patient load,” and “satisfaction with 3 core tasks,” which were “finding priority patients,” “reviewing data charts,” and “interacting with patients.”
The valid responses revealed 3 signals, which answered three questions: “where to dig deeper, which feature flows hold the evidence, and how to choose interviewees to cover the full picture.”
Signal 1
Of the 3 core moments, "finding priority patients" had the weakest satisfaction

The easiest of the three to overlook.
Signal 2
Different roles use very different ways to find patients, with no shared pattern

The easiest of the three to overlook.
Signal 3
Different roles carry very different patient loads and work in very different ways

This is why "finding priority patients" can't be solved from just one role's perspective.
Based on these 3 signals, I further narrowed down candidates for the field interviews.
I prioritized clinicians whose satisfaction ranged from “neutral” to “agree,” covering different care roles and patient loads to understand a wider range of real-world use.
During the interviews, we found that some clinic types offered limited insights, so we expanded the candidate pool from 8 to 12.

From Opportunities to a UX Backlog
I wrote each observation from the 12 interviews on its own sticky note, then grouped them by theme into 41 improvement opportunities and aligned priorities with the team.
View the full breakdown in Figma
“41 opportunities, grouped by mentions, clinic coverage, and effort”
➊ Quick Wins (add to this round's backlog)
➋ Worth Deeper Research: Feature Discoverability
The most-mentioned issue across the list was feature discoverability (6 mentions, 4 clinics).
As clinicians put it: “We need these features, but we don’t know where they are or how to use them.” The team saw value in the direction, but the scope was large, so we’ll first define how far the Key Feature Tour should go, then validate it in phases.
Prioritized HMW opportunities from the interview findings
Here are a few cases from the 41 opportunities: quick wins that have shipped, completed design proposals, and directions worth exploring further.
➊ Clearer Feature Semantics: From Rarely Clicked to Willing to Try
Interviews revealed that many users couldn’t tell what the icons did, so they never clicked them. We reworked the icons, labels, and layout to make each feature’s purpose clearer while taking up less screen space.

➋ Simplifying Multi-day Trends: Validate Small, Then Invest Big
CGM is continuous monitoring and is meant to be read over 14 days, but 14 days of overlapping lines were hard to read. In the short term, we tested the smallest change, defaulting to 7 days; in the long term, from a design perspective, a different visualization would be the better fix, so we explored a CGM trend grid that splits each day into its own cell to reduce overlap. The team saw the direction as worth pursuing, but given the development cost, we shipped the 7-day default first.

➌ In-place Replies: Quote Photos Directly While Viewing
Diet is central for clinicians tracking diabetes patients, and a single diary entry often holds several meal photos; the most common scenario is giving advice on a specific food in a specific meal. But after enlarging a photo, clinicians couldn’t quote it in a reply, had to exit the view first, and found it hard to point to which photo they meant. So we added Quote reply to the enlarged view, letting clinicians quote the exact photo they’re viewing without breaking the flow. This is a design proposal that hasn’t shipped yet, but clinicians responded positively after seeing the prototype.
hese are three of the 41 improvement opportunities; the rest, with their issues, mention counts, and resolutions, are documented in the full opportunity list.
Impact & Validation
📊 Quantitative Results
After the label update, AI Analysis events increased by 50% and active users by 17%.
The patient notes button declined: events −22%, active users −10%. Users fell by only 10%, mainly because each person opened it less often (about 19 → 16 times); AI Analysis grew over the same period, so overall usage didn’t drop. Instead, the clearer icons led to fewer misclicks.
Due to limited event tracking and engineering resources, other improvements were evaluated mainly through qualitative feedback. Overall, clinicians found the updated experience clearer and easier to use, while also identifying opportunities for further improvement.
bottom of page