(Illustration by Robert Nuebecker) 

Practitioners in the field of humanitarian aid increasingly acknowledge that conventional approaches to conflict intervention have not produced satisfactory results—despite, in some cases, the expenditure of vast sums of money. There are numerous reasons for this failure, but the most salient one is that programs that rely on grand planning and rigid implementation work only under very specific conditions, almost none of which apply to conflict zones. By definition, conflict intervention takes place in fast-changing, volatile environments in which it’s impossible to know how a given intervention will unfold.

At the same time, there is a growing awareness that other, more efficient techniques exist for making decisions and designing interventions in complex environments. Social service organizations have started adopting experimental methods—also known as iterative or lean methods—to address enduring social problems like poverty, illiteracy, and disease transmission. In this kind of approach, practitioners quickly launch small-scale pilot tests of program ideas, and then cancel programs that are ineffective and work to enhance programs that show promise.

These techniques are also gaining traction within traditional international development organizations. Researchers at the Center for International Development at Harvard University, for example, have created an approach known as problem-driven iterative adaptation. And professionals at Mercy Corps have developed a model that they call adaptive management. As yet, however, few practitioners have effectively incorporated these techniques into conflict management or into postconflict interventions.

Unfortunately, heightened awareness of the limits of conventional programming in unstable environments has not led to the adoption of truly innovative approaches. The current approach reinforces a fruitless cycle: International NGOs draw support from donor agencies to provide services that sometimes mitigate short-term conflict effects but do little either to resolve conflicts or to create a foundation for achieving long-term development goals. Any program that is crafted from afar, that locks in program strategies ahead of time, and that is implemented largely by external actors who have a limited understanding of beneficiaries’ real lives is unlikely to deliver effective results.

Instead, organizations that work in conflict zones need to cultivate an approach that uses lean or iterative techniques to identify solutions that can take root locally and that demonstrate a potential to become self-sustaining. They need to adopt processes that replace grand planning with small experiments—processes that focus on incremental performance improvements rather than predetermined milestones. At my organization, Bancroft Global Development, we are working to develop and implement such efforts.

The Need for Flexibility

Over the past few decades, the business world has seen the emergence of several process and product improvement platforms. Examples include human-centered design, a product innovation method developed by the design firm IDEO, and lean experimentation, an entrepreneurship method that originated in Silicon Valley. Both of those platforms emphasize experimentation and rapid iteration, strong feedback loops that facilitate early and continuous engagement with end users, and the use of minimally designed prototypes to test products or processes. Taken together, these elements enable developers to discover and build on what works, to jettison what does not work, and, when necessary, to “fail fast”—before they have expended significant resources or large amounts of time on a project. In contrast with conventional methods for delivering international aid, which rely on elaborate strategic plans and regimented implementation, these techniques are well suited to pursuing interventions in unpredictable settings.

To be sure, there are significant differences between the goals and responsibilities of a typical business enterprise and those of an organization that implements conflict or postconflict interventions. A business, for example, can measure its success by calculating its market share or by analyzing its profit-and-loss statement. In a conflict intervention, however, practitioners generally evaluate their work by tracking the number of lives improved (or saved), the progress of policy reform, and other forms of impact that are difficult to measure or to associate with a single causal factor. Even so, the core reason to adopt iterative problem solving—the wish to develop and deliver products or services that fill a social gap or serve a human need—applies at least as well to conflict intervention as it does to business operations.

Our work at Bancroft provides an example of how organizations can apply iterative techniques to conflict zone intervention. We specialize in developing local capacity for managing conflict-to-postconflict transitions. In 2008, during a period of intense civil conflict in Somalia, we began collaborating with the African Union Mission to Somalia (AMISOM) and the federal government in that country. Along with those entities, we have worked to enhance the capacity of Somali institutions to provide security and public services, to reduce the threat posed by the militant group Al Shabaab, and to facilitate humanitarian assistance.

A conventional approach to this intervention might have identified one or two (or five or six) objectives, translated those objectives into individual programs, and sought funding to undertake these programs for a specific duration. Instead, we started by placing a small number of expert mentors in the country to work alongside counterparts from AMISOM and the Somali government. These mentors then developed specific strategies and activities in collaboration with those institutional partners.

Initial efforts by the mentors centered on enhancing peacekeeper skills—tactics for countering IEDs (improvised explosive devices), for minimizing civilian harm, and so forth. Then, in the first year of this collaboration, the mentors and their counterparts began to adopt approaches that were not just reactive but also preventive. They incorporated the use of explosive-detecting dogs, for example. Later, after several foreign governments and multilateral institutions reestablished aid efforts to Somalia, providing policy support to the Somali government became a top priority for our mentors. At each of these stages, our team of mentors learned from their experience in the country and adapted their methods accordingly.

In subsequent years, AMISOM and the Somali government expanded their reach and increased their level of engagement in Somalia, and our work with these partners continued to evolve. We now had to provide not only peacekeeping support but also assistance to populations in newly recovered territories. We began to prioritize logistics, communications, engineering, and medical capabilities. Because of our commitment to experimentation and iterative problem solving, Bancroft and its partners were able to streamline, redirect, or expand the deployment of personnel and resources in real time as facts on the ground changed. Crucially, we were able to complete many such adjustments over the span of days or weeks rather than months or years.

The Lesson of Failure

Mentoring is highly amenable to iterative programming because it supports a process of self-directed learning on the part of an individual protégé or a host organization: As the individual or members of the organization gain new skills by working with a mentor, they can take on more ambitious projects; as they discern new problems to address, they can work with the mentor to develop additional expertise. In a conflict zone, that kind of adaptive flexibility can be a crucial asset.

But other efforts to deliver aid in conflict or postconflict settings—direct service programs, for example—would also benefit from the use of lean or iterative techniques. Practitioners can adopt such techniques by following a few key principles:

  • Start with a small team to test the design and evaluate the effectiveness of a new program.
  • Continuously solicit input from end users (and not just from donors) during program implementation.
  • Whenever possible, incorporate existing local resources and systems into a program before importing outside resources and systems.

Donors, meanwhile, should accommodate the need for this kind of programmatic flexibility by (for example) funding experimental pilot programs. International development professionals are well aware of the many counterproductive pressures that affect their field: the expectation that programs will deliver results on election-cycle timelines, the need to appease hidebound spending authorities, the tendency to focus on easily measured transactional activities rather than high-impact transformational activities.

Chief among these pressures may be the taboo against failure. Because delivering lasting international development solutions is inherently difficult, every organization in this field experiences failure at some point. Silence on this topic is equally universal. People are quick to tout nebulous project achievements and slow to acknowledge glaring signs that a program is ineffective or inefficient. It’s easy to understand why: Donor agencies routinely base their funding decisions on whether recipients have a reputation for “proven” success. And donors accept these dubious claims of success because they face pressure to disburse funds at a rate that will justify their own budgets.

Iterative practices can build a habit of learning through experimentation, reduce the stigma that attaches to failure, and empower practitioners to discuss adverse results openly and rationally. By adopting iterative problem-solving methods, organizations can do away with the false front of “success” and increase the tolerance for productive risk taking. Experimentation, after all, is a process in which successive failures generate knowledge that ultimately leads to real successes.

Tracker Pixel for Entry