While successful for years, the paper-based BRAC Poverty Reduction Program needed a scalable white-label PAAS update which optimized analysis, modeling and cohort success rates.
Reduce participant intervention times by ~20%
Improve cohort engagement by ~15%
Attract new Investors by ~25%
Add ~30% more eligible participants
It was regularly difficult to get succinct deliverables and due dates which made planning and execution impossible. Language barriers sometimes hindered rapid understanding of contexts, problems and solutions. Time zone differences made timely and actionable feedback difficult.
I conducted field research with product manager on participants in rural Bangladesh to validate cohort population numbers, program completion times and average success rates.
I toured small villages to understand the cohort journey from beginning to end, studying all phases of the 2 year journey, with particular attention placed on how mobile devices would be used to collect data.
I reviewed summary of research with product manager and engineering lead to set measurable success targets for effective features.
I photographed all paper based assets to document previous data-gathering and program administration methods and tools.
All research was conducted to align with larger business goals of program optimization, streamlining goal flows, reducing annual program costs, and building a database to enable historical analysis and future modeling. Particular attention was placed on understand the experience of the field workers who literally go site to site to use mobile devices to monitor participants.
It was hot. 110 degrees in the shade. It was only the single PM and myself, and I did not speak Bengali. Luckily the PM did, but often after long conversations the PM had with participants in Bengali, he had to summarize and translate for me, and a lot I'm sure was lost in translation.
I took tons of photos to document all paper assets used in managing the program participant experience. I reviewed these with the PM in the evenings and back in San Francisco to strategize UX asset delivery, determine overall project scope and keep stakeholders aligned on progress.
We had abundant documentation with pictures, interviews, and anecdotal evidence.
We had a solid knowledge of how mobile devices would be used.
During subsequent wireframing and hand sketching we were able to plan the project...
During the research process we were also doing a lot of planning. We were constantly distilling our findings down into wireframes, whiteboards and later prototypes.
The planning was done to align to all higher level goals for improving cohort success rates, but also to align to known and desired engineering resources. As this program relied on having field workers go from village to village, we were focused on a mobile first experience.
Time is always a challenge. We were under pressure to deliver timely results, and time zone and language barriers contributed to misalignments, delays and product definition churn.
Copious white boarding in the office with stakeholders, engineers, designers, product management. We frequently sketched out future features to facilitate rapid feedback.
A key outcome for the planning process was a comprehensive experience map geared toward helping the program administrators gather data in the field with mobile devices.
This was an absolutely essential asset used to cohere many abstract concepts down into a tangible and actionable visual understanding.
As the planning phase was progressing, in tandem I began to visualize what the different platform components would need to succeed.
Using the experience map as a guide, I began to iterate on wireframes aligned to optimizing cohort success.
These wireframes were specific to all mobile and desktop views, as different program administrators would be using different devices at different phases of the journey.
I color coded the relevant pages aligned to the swim lanes of the experience map.
Basic goal was to start visualizing the final result and share all progress with engineering. This ensured all the teams cohered around the same ideas.
Focus was placed on the mobile experience of field workers who were collecting the data, but we knew the data would get re-presented for analysis and modeling on larger devices later.
Again, time. There was so much work to be done, and goalposts kept shifting, ensuring team alignment was often difficult and stressful.
As I was the chief UX person in the project, I had to work quickly. This was a Sketch shop, so I used Sketch as the primary design tool for all wire-frames. Axure Pro was used for all interactive prototypes. The team was tight-knit, so myself, the PM, designer, and engineers worked closely and met daily to discuss progress, blockers and any schedule adjustments.
I built prototypes of both mobile and desktop experiences, which grounded the team in a mutual vision of the final product.
The current successful project continues to drive a 60% reduction in demographic poverty.
The final white label result is being used in different social impact contexts even outside of poverty reduction.
Collected data enables transformative budgeting, modeling and error correction.
Average reduction in program completion time ~30%
Improved cohort engagement and retention ~20%
Net new Investors and business growth ~20%
Additional eligible participant improvement ~10%