meals served
kids fed every school day
average attendance, unannounced count
of malnutrition cases recovered
How we count
The number we care about is attendance on a boring Tuesday.
It's easy to produce a good figure on the day a donor visits. So we don't use those days. Attendance is taken by a partner staff member on a date the school isn't told in advance, once a term, and that's the figure we publish.
| Measure | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Meals served | 3,700,000 | 4,300,000 | 4,900,000 |
| Children fed daily | 19,400 | 22,800 | 26,400 |
| Average attendance, unannounced count | 71% | 78% | 84% |
| Girls as share of children fed | 51% | 55% | 58% |
| Food purchased in-district | 51% | 64% | 76% |
| Malnutrition recovery rate | 83% | 86% | 89% |
| Food cost per plate | 27¢ | 26¢ | 25¢ |
Full methodology, including how we handle schools that leave the program mid-year, is in the annual report.
The honest caveat
We can't prove all of this was us.
Attendance in our districts rose while we were feeding kids there. Some of that is the meals. Some is government policy, some is rainfall, some is a road getting graded. We don't run randomized controlled trials — we're too small, and it would cost money we'd rather spend on maize.
What we can say with confidence is narrower and duller: 4.9 million plates were served, logged daily and reconciled against grain receipts; 3,120 children completed malnutrition treatment; 9,400 rations went out. Those are counted, receipted and checkable. The rest is inference, and we label it as such.
If you want the well-evidenced version of the general claim, the World Food Programme and the Copenhagen Consensus have both published extensively on school feeding. We'd rather point you at their evidence than overstate ours.
See the financials behind these numbersPick an amount. See exactly what it buys.
Every gift is receipted, pooled monthly, and reported back to you with real numbers.