CASE STUDY 01
Measuring
digital service
Performance
A performance dashboard that stopped showing data - and started telling the truth behind it.
AI-ASSISTED RESEARCH SYNTHESIS
CLIENT
Govt. of Alberta
DOMAIN
Public Sector . Service Design
MY ROLE
CORE METHODS
Senior Product Designer
Usability testing . Research synthesis . Hi-fi design
IMPACT
ADOPTION
EXECUTIVE CONFIDENCE
COGNITIVE LOAD
OVERVIEW
Where every design decision
has consequences.
Working within the public sector means navigating complex governance environments, cross-ministry dependencies, and high-impact decision pathways. Even minor design changes can have wide-ranging effects, making thoughtful planning and stakeholder alignment essential.
The challenge: a dashboard that needed to do more than display numbers. It had to accurately reflect how every service under every ministry was performing, reveal the true story behind each metric, and show how effectively public funds were being utilized.
THE PROBLEM
A tool
nobody trusted.
The primary challenge was low user engagement and adoption. Each quarter, teams were required to update the dashboard with new performance metrics - some standardized, others unique to individual services. To users, it felt like yet another administrative burden piled on top of already heavy workload.
" When metrics didn't meet expectations, users had no place to add context - leaving executives with an incomplete, often misleading picture."
Compounding the issue: certain metrics were measured only once, yet the dashboard displayed them as permanently stagnant - because they was no way to mark a target as complete. Over time this created a cycle of frustration. Users dreaded returning. Updates were delayed. And when they finally did come back, they couldn't remember how to navigate the tool.
The result: a tool that was neither trusted nor effectively used by the people it was built for.
RESEARCH AND DISCOVERY
Listening before
designing.
My first step was to deeply understand the product, the people using it and the operational realities behind their frustrations. I recommended starting with usability tests - segmenting users into meaningful groups so each session could surface distinct and relevant insight.
GROUP 01
GROUP 02
GROUP 03
Frequent Users
Lapsed Users
New Users
Power users who knew the system's workarounds and could articulate exactly where it broke down.
Had access but rarely engaged - their avoidance was most telling signals about adoption barriers.
Fresh eyes that revealed onboarding failures invisible to anyone already familiar with the tool.
For many participants, it was the first time they felt truly heard. All sessions were moderated with detailed observation notes captured throughout.
USABILITY TEST SYNTHESIS - MIRO
AI IN THIS PROJECT
From raw notes to clear themes - faster.
After completing usability sessions, I used AI to hep process and cluster of qualitative feedback. Interview transcripts and and observation notes were fed into an AI synthesis tool to surface recurring language patterns and behavioral themes - work that previously took days of manual affinity mapping compressed into focused hours. The output wasn't the insight, my output was. AI just cleared the noise.
TOOLS USED: CHATGPT . NOTEBOOKLM
POST USABILITY TEST WORKSHOP SYNTHESIS
DESIGN SOLUTIONS
Targeted enhancements,
not a redesign.
1
Monthly Metric Update
Greater granularity in how performance is tracked - shifting from quarterly to monthly inputs to reduce cognitive load and improve accuracy.
2
Context Field
A dedicated space for users to explain unexpected performance trends - so executives see the story not just the trend.
3
Cost Data Integration
4
Setup Wizard
Financial visibility added per service - enabling leadership to connect investment decisions to actual outcomes.
An always-accessible onboarding tool so users could re-learn the dashboard without hunting through documentation.
5
Status and End Dates
Metrics could now be marked complete - eliminating the false "no progress" signal that had been eroding trust in the data.
6
External Link Integration
Direct link to supporting documents within the tool - reducing the friction of context switching mid-task.
HIGH FIDELITY DESIGNS
The screens
that shipped.
THE IMPACT
What actually
changed.
These weren't incremental polish updates. Each enhancement directly closed a gap between how the systems worked and what users needed - resulting in a tool that was used more, trusted more, and made better decisions possible.
Improved Performance Visibility
Monthly inputs replaced quarterly aggregation - giving terms and leadership a far more granular, accurate view of service health.
Reduced Cognitive Load
Smaller, more frequent updates eliminated the mental overhead of recalling months of data in a single session.
Increased Trust & Engagement
Users who felt heard during research became advocates. Feedback loops between teams and stakeholders became healthier and more constructive.
Higher Platform Adoption
Clearer workflows and more honest data storytelling gave users confidence that the dashboard reflected their work accurately.
Better Executive Decisions
Context fields and target timelines gave leadership the narrative behind the numbers - not just the numbers themselves.
Financial Clarity
Cost data per service enabled leadership to connect investment decisions to service outcomes for the first time.
NEXT CASE STUDY
IMPROVING MINING OPERATIONS @NUTRIEN