Our Recommendation
DMI is a top‑tier, rigorously evaluated behavior‑change organization that meets and often exceeds our bar on intermediate and ultimate outcome evidence, including a large‑scale RCT with child‑mortality effects. For a funder using our framework, DMI is a strong candidate for substantial support: their theory of change is tightly linked to specific negative consequences (care‑seeking and child deaths), their intervention is focused, scalable and cost‑effective, and they already have gold‑standard counterfactual evidence, with clear scope to deepen measurement on nutrition, gender norms and economic effects.
DMI's Fierce Certification score is 120/100 based on our criteria:
✔ Has Ultimate Outcome Goals (50 pts)
✔ Measures Intermediate Outcomes (10 pts)
✔ Measures Ultimate Outcomes (15 pts)
✔ Shows Continual Learning & Adaptation (25 pts)
✔ Measures Intermediate Counterfactual (10 pts)
✔ Measures Ultimate Counterfactual (10 pts)
The Social Problem
DMI is addressing high, preventable burdens of illness and death in low‑income countries, especially among young children. Despite existing health services in many countries, parents often do not recognize danger signs in malaria, diarrhea or pneumonia or seek care quickly, and women under‑utilize maternal and reproductive health services, leading to avoidable mortality and morbidity. Weak health knowledge, harmful beliefs and under‑used health systems collectively slow progress toward national health goals and SDG 3.
The Solution
Their solution is intensive, evidence‑based mass‑media campaigns (radio, TV, mobile and other channels) that use engaging stories and spots to promote specific, life‑saving behaviors. DMI applies scientific modeling and formative research to identify the behaviors that will save the most lives (such as prompt care‑seeking for child fever, diarrhea or pneumonia), designs creative content to motivate those actions, broadcasts it frequently enough to shift norms, and evaluates impact through rigorous trials. The theory of change is that by changing knowledge, attitudes and norms around key health behaviours at massive scale, they can substantially increase uptake of proven health services and thereby reduce deaths and improve health outcomes.
Key Outputs
Key outputs that contextualize DMI’s work:
- Scale and reach: DMI’s campaigns reach tens of millions of people; one external evaluation notes over 90 million people reached with health campaigns in sub‑Saharan Africa in a given year.
- Childhood‑survival RCT: a 5‑year RCT in West Africa tested an intensive radio campaign and demonstrated large increases in treatment‑seeking for child illnesses and a 6.3% reduction in under‑5 mortality.
- Range of health topics: they now work across child survival, reproductive health, nutrition, hygiene/sanitation and early childhood development.
- Recognition by evaluators: organizations such as GiveWell, The Life You Can Save, Founders Pledge and Mulago have highlighted DMI as a highly cost‑effective, evidence‑backed organization.
These outputs show an organization that is both research‑intensive and delivery‑at‑scale, anchored around a few core behaviour‑change levers.
Key Intermediate Outcomes
Intermediate outcomes with strong measurement (and counterfactuals in key cases):
- Care‑seeking for child illness: the West Africa RCT found 56% more children taken for treatment for malaria, 73% more for diarrhea and 39% more for pneumonia in intervention vs control areas, directly addressing low care‑seeking.
- Maternal and reproductive health behaviors: evaluations show increased antenatal care attendance, facility deliveries, and uptake of modern contraceptives following DMI’s campaigns, though designs are often quasi‑experimental rather than full RCTs.
- Knowledge, attitudes and hygiene practices: post‑campaign surveys report higher knowledge of disease‑prevention behaviors, improved infant‑feeding practices and more consistent hand‑washing and sanitation behaviors among exposed populations.
All of these changes in behavior and resource use represent vital steps toward improved health but are distinct from the health‑status outcomes themselves.
Key Ultimate Outcomes
On ultimate outcomes, DMI is unusually strong:
- Child mortality reduction: the flagship RCT estimated a 6.3% reduction in under‑5 mortality in intervention areas as a result of DMI’s child‑survival campaign. This is a direct, counterfactual estimate of an ultimate outcome.
- Mortality and fertility in other campaigns: external summaries note additional mortality and fertility reductions associated with DMI’s work, though these are described with less methodological detail than the main trial and often as model‑based estimates.
Other ultimate outcomes (nutrition status, psychological wellbeing, productivity, progress on national health goals) are more often inferred than directly measured. DMI’s core claim. reducing child deaths through behavior change, is backed by unusually rigorous evidence for a media‑based organization.
Continual Learning & Adaptation
DMI exemplifies a learning organization aligned with our four‑step cycle:
- Evidence at the centre of strategy: they describe themselves as combining “science and creativity,” with a priority to “continue to generate ground‑breaking research” and then “take proven strategies to scale,” signaling a feedback loop from evidence to implementation.
- Use of modeling and trials to refine focus: DMI uses modeling to identify which behaviors will save the most lives per dollar, then tests campaigns via RCTs and quasi‑experimental evaluations, adjusting content and intensity in response to results.
- External vetting reinforcing internal learning: close engagement with evaluators like GiveWell, Founders Pledge, Mulago and The Life You Can Save both reflects and reinforces a culture of measurement, challenge and iteration.
Within our framework, DMI is strong across all four steps: a clearly articulated, problem‑grounded theory of change; an intervention tightly mapped to that theory; robust measurement of both intermediate and ultimate outcomes with counterfactuals; and an explicit commitment to learning and scaling based on what the evidence shows.
Written by
