Cohort tool
An analytics tool that helps public health professionals compare patient cohorts over time using granular data.
Before
After
Project Challenge
​
Cohort analytics is a type of analytics that helps public health professionals compare individual patient cohorts over time; each cohort shares a common characteristic such as time of outreach, infection date, or treatment
regimen that can be defined very granularly on any available data.
By analyzing cohorts, M&E teams can assess cascades of care across cohorts and tweak features to their interventions to more granularly determine ways to increase metrics such as treatment retention, engagement with the program, viral load etc. These types of analyses are key to understanding the key risk factors for transmission and the medical outcomes of subsets of the affected population.
​
The team had previously built this tool based on inspiration from a marketing tool. The mental model of the user wasn't taken into account. Mixed with the inherent complexity of the tool, many users required active hand-holding from our customer success team. How might we build a tool that allows users to feel confident in calculating complex cohorts?
Involvement​
​
Design research
UX/UI Design​
​
Client​
​
Zenysis
​
timeline​
​
Feb - Apr 2020
​
teamates​
​
Quentin - Product Manager
Stephen - Principal Engineer
David - Senior Engineer
Project Solution
​
Based on a round of research with users, progressively iterate on the current analytics tool to allow users to efficiently, successfully and confidently take an analytical problem and complete the task end to end.
ux challenges
How might we redesign the cohort tool to allow users to efficiently, successfully and confidently take an analytical problem and complete the task end to end?
Business Value
Reduce the amount of time our customer success team spends hand-holding
Customer Value
Make cohort analytics tool a complete end-to-end tool so users do not have to rely on disconnected tools like Excel or our customer success team.
KPIs for success
-
Increased program success in governments we work with
-
Reduced amount of time customer success team spends with users
Research approach
Usability Study
We conducted 5 usability tests with two sets of personas: (1) Program Data Managers, and (2) Customer Success team members. The Program Data Managers were our primary persona. They work in the governments we collaborate with and are the technical data professionals who are responsible for analyzing program success. We also conducted tests with our internal customer success team, as they are responsible for driving success with our primary users.
Finding patterns
The product manager and I synthesized the results and boiled them down to two main problems
ideate, prioritize, deliver
Design thinking workshop
The engineers working on this project were involved from the beginning. Instead of prioritizing problems and ideating on solutions in a vacuum, I drove a design thinking workshop that involved my colleagues.
ideate, Group and vote
As a group we ideated on possible solutions then grouped these solutions into similar buckets. Taking both effort and value into account (which was the advantaged of having engineers included), we dot voted on our favorite solutions.
choose and create
Each person took their favorite solution and quickly prototyped via pen and paper. This was at the beginning of COVID so we conducted everything via Miro and uploaded our solutions digitally.
product solutions
Based on the design thinking workshop, a few potential solutions rose to the top that together would solve for the two overarching problems. We were faced with a limited bandwidth and faced the tough decision of dropping rethinking master logic.
.png)
The final product reflected these three buckets of solutions.
