What International Development Can Learn from Agile Software Development

What International Development Can Learn from Agile Software Development

By Andrew Howe and Mary Grace Reich | Program Manager, Growth Operations Lead

Have you ever collected data that was not fully utilized—maybe made it to a final report but never informed activities? Is the data you just collected now illuminating what you only wish you had asked instead?

The international development community is abuzz with agile, adaptive or iterative management because adaptive management produces better results. From USAID’s approach to Collaborating, Learning, and Adapting (CLA) to the Doing Development Differently Manifesto, the community has committed to managing adaptively for years. However, in practice, resources, structures and processes have inhibited international development practitioners from optimizing agile management.

Agile management has been tried, tested and implemented with increasing effectiveness in software development. In software development, agile management techniques allow teams to advance quickly on proximal goals. Teams typically focus on two-week “sprints” in which a set amount of work is selected. At the end of two weeks, teams gather to review progress and issues as well as select the work to accomplish in the next “sprint” with any course corrections deemed necessary. This cycle of planning, implementing, evaluating and adapting closely follows the frameworks for adaptive management to which international development aspires.

“Failing fast,” a core tenet of agile techniques and software development cannot be applied as loosely in international development. Best practices in international development using agile techniques must be carefully evaluated to ensure no harm is done. That said, agile-inspired management can help program implementers monitor progress, as well as shed light on assumptions and risks. 

Adapting quickly requires learning quickly to inform those adjustments. Software development inherently includes a quantitative feedback mechanism—log files of events, failures and other built-in checks. For the diverse and dynamic environments of international development programming, feedback loops are more difficult to implement and respond to. But the urgency to improve outcomes is even more intense when opportunities to better the human lives most in need are at stake.

At Premise, we employ agile-inspired tools and techniques to help our partners collect quality data and generate efficiencies. Combined with citizen-led science and a responsive feedback mechanism, Premise can iterate in several ways. Three general examples include 1) tools to collect data; 2) analytics; and, 3) insights utilized for making decisions. 

Iterative Tools for Data Collection

For an ongoing project, Premise sought to understand the cost of a new at-home, self-injectable contraceptive to help our partner understand barriers women may face in accessing the product. Upon collecting pilot data, enough outliers warranted closer inspection. Through visiting a couple of facilities to gain a better understanding of the costs reported, Premise learned that the reported costs were not outlier data. The high variability in costs represented the product as well as consultation and registration, which are often bundled together. This helped Premise to quickly adapt the line of inquiry with the client and the instrument—a task on the Premise app—to generate more accurate data and insight into barriers and opportunities in accessing the product.

Iterative Tools for Analyses

More and faster data requires resources to efficiently digest it. In a recent project piloting experimental indicators of quality of life and safety across sub-city localities, three iterations of data delivery enabled greater efficiency in communicating results. For the initial analysis, Premise employed a visualization of descriptive statistics to surface the necessary nuance to evaluate the indicators. Subsequently, an initial map of the data tested analyzing the most compelling indicators geospatially. After engaging with both formats, indices which summarized correlating indicators informed a second map that could efficiently communicate results. Data analysts, technical experts and iterative tools were all critical components to achieving a visualization which enabled other stakeholders to more efficiently learn from the results.

Iterative insights

Programs are often large enough in geographic scope that blanket implementation lacks timely, efficient action. Dynamic engagement with the data collected and analytic tools developed can help increase effectiveness. In Cali, Colombia, Premise worked with the city secretary of health to monitor mosquito larvae that carries the Zika and Dengue viruses. Based on citizen observations of larvae throughout various neighborhoods, we helped create heat maps to assess risk that the city used to target their efforts. The result was continued risk mitigation with greater efficiencies.

These are three small examples of how Premise is working to enable an efficient feedback mechanism in international development, so our partners can learn quickly, adapt quickly and optimize results. 


If you are interested in learning more about how to incorporate agile management into your programs, check out more on our website at www.premise.com/international-development or emailing andrew@premise.com

About Andrew Howe and Mary Grace Reich

Andrew Howe manages the Bill & Melinda Gates Foundation and USAID projects at Premise Data. He first joined Premise Data in the engineering department before transitioning to manage international development projects marrying his background in research and professional activities in Africa.


Mary Grace Reich is a Growth Operations Lead at Premise, where she works on building tasks and engaging the Premise Contributor Network to complete them for our international development projects. Before Premise, she worked in international development program management on initiatives ranging from public access technology and digital literacy to digital security.