My Responsibilities

• Brainstorming solutions in collaboration with UX researcher
• Building design system
• Building low fidelity and high fidelity wireframes
• Presenting solutions to stakeholders and receiving and incorporating feedback
  
• Developer handoff

Setting the Stage

Pakistan is one of the two remaining polio-endemic countries in the world, alongside Afghanistan. To help eradicate this disease from the world, the Bill and Melinda Gates Foundation (BMGF), paired up with the National Emergency Operations Centre (NEOC) to launch Pakistan’s National Emergency Plan for Polio Eradication.At the centre of Pakistan’s polio vaccination efforts is the Geographical Coverage Support System (GCSS) dashboard, which uses real-time data from health workers in the field to monitor and control vaccination campaigns, in addition to identifying visited and missed areas.

The Challenge

Redesign the dashboard to address evolving needs and fix usability issues so users can receive live feedback on geographical coverage of the polio campaign

Our Users

• GCSS Team
  The team was responsible for identifying and then covering areas where children had not been vaccinated
• Area-in-charge
  Each GCSS team was categorised on a provincial and district level. The supervisor or Area-in-charge was responsible for monitoring and ensuring smooth running of the campaign
• Investors
  Since the program was funded by BMGF, they needed to gauge the success of the  polio program through this dashboard

Key Use Cases

We were approached to add a few missing features to improve the functionality of the dashboard. We conducted a design sprint with representatives of the NEOC to pinpoint the use cases which had the highest priority. We also had meetings with BMGF investors to identify their needs.

GCSS Team and Area-in-charge needed to be able to

Use Filters to show granularity of data

See missed areas and other location markers

See KPIs clearly

Use NEOC credentials to login

Access key features on mobile

BMGF Representatives needed to be able to

Estimate the success of each campaign by province, district, tehsil, and union council level

Ensure consistency of nomenclature across the tool and on ground campaign

Make sure KPIs clearly communicate the accuracy level

Show tracking on an individual level

Separate out intra-campaign, post-campaign, and compare campaign dashboards

Once we had the key use cases defined, we conducted an informal audit of the existing dashboard. We found that other than the missing features, the dashboard had some glaring usability issues. This is where I suggested we should shoot for a complete overhaul rather than adding features to an already clunky foundation. With this approach, we could make major changes to the layout and user flows rather than minor iterations.

How we did it

Brainstorming, Wireframing, Design System, and Hi-fi Mockups

Initially we started out with minor fixes to the dashboard to address our users’ needs. However, as we combed through the existing dashboard, we found many errors and minor UX issues. Date selection was unintuitive, some filters were in the wrong place, and some things that could be buttons were drop down menus, and there were major hierarchy issues as well where the most important information (KPIs) were hidden away.

For the low fidelity wireframes we brought in other designers to brainstorm solutions together on how the dashboard could look like. At this stage, we were thinking about the prioritisation of information and which overall layout would work best. We combined the best ideas to move forward with a single approach.

We decided to adapt the Ant Design System since it met most of our needs and saved us valuable time. We supplemented it with our custom components where needed.

Once we had the design system in place, we built out the dashboard and iterated on a daily basis. We received feedback from stakeholders to ensure user needs were met and our solutions were easy to understand and adopt.

What we made

We built the new design to solve the usability issues and cater to the new requirements. Iteration was key here and involving the stakeholder for feedback constantly throughout the process was essential. The key problems and solutions are outlined below.

Usability Study Changes

• We found out through our study that the KPIs were more important than the map in the hierarchy and shifted them up
• The icons for missed areas were difficult to distinguish according to confidence level and had to be changed
• Home section and onboarding screens were deemed unnecessary and removed
• Users requested a feature where they could easily have a high level view of KPIs which led us to the control room view

Did it work?

In just one month, the dashboard was used to identify 12,715 missed children across 551 Missed Areas. Of these, 5,357 children were successfully vaccinated. During the span of two years, the GCSS dashboard has identified 6,643 missed areas and vaccinated missed children across those saving thousands of children from polio. Data was not available from the previous design of the dashboard for a comparative analysis. However, our users had a couple of things to say.

Was it better?

Key Learnings

• Communicate often: It is natural to make assumptions to fill gaps in our understanding. However, there is no room for assumptions when you’re working on a tool responsible for saving millions of lives. Solid and frequent communication was key while working with large cross-functional teams to avoid details getting lost in complicated jargon.
• Share frequently: When your design is solving for multiple stakeholders, as well as business needs, it is always helpful to show what you’re doing and the next steps you’ll be taking. This helps validate or pivot as needed and saves time and effort in the long run. Having weekly and daily check-ins helped us maintain a constant feedback loop.
• The design is never final: Design is a constant iterative process. Therefore it is important to stay nimble, keenly listen to feedback, and be willing to pivot when needed. There were times during this project we were making changes and getting feedback daily.
• Document your work: We maintained a process book throughout the whole project and it helped us come back to retrace our steps whenever we felt stuck. Additionally, it helped us see how we revamped the dashboard from start to finish. For a fairly large project, you begin to forget the details moving forward, so it’s always helpful to have them in one place.