Case study: one MEAL platform for 12 NGO partners in 8 countries
A German humanitarian donor needed one reliable view of a regional crisis response run by 12 local and regional NGO partners, through 13+ projects in eight countries across East and Southern Africa. In a four-month engagement we standardised the framework, built the platform and trained the teams who now run it in all eight countries.
At a glance
- countries, live
- 8
- NGO partners (11 local, 1 regional)
- 12
- active projects
- 13+
- partner data tools reviewed
- 84
- standard indicators
- 74
- standard data collection tools
- 13
- response pillars
- 4
- thematic areas
- 4
- training participant-sessions
- 50+
Countries: Ethiopia, Kenya, Somalia, South Sudan, Madagascar, Malawi, Mozambique, Zimbabwe.
Four thematic areas
- Food Security & Nutrition
- Livelihoods & Markets
- DRR & Climate Change Impact Mitigation
- Protection & Psychosocial Support
Four pillars, one crisis response continuum
- 01Anticipatory Humanitarian Action
- 02Emergency Response
- 03Recovery
- 04Resilience Building
The challenge
Each partner tracked its results in its own Excel files, which were then consolidated by hand. Data quality varied from partner to partner, nothing was available in real time, there was no version control and no shared rules for how data should be governed.
Even the basics differed: some partners split ages into four bands and others into two, some counted persons with disabilities inside their totals and others outside, and household figures were converted to people with one fixed multiplier for every country.
What we built
- 01
Desk review and decision register
We reviewed the theory of change, logframe, indicator reference sheets, MEAL matrix, indicator tracking table and all 84 partner data collection tools, then resolved every inconsistency through a documented decision register approved by the donor before any code was written.
- 02
One standard framework
A single goal, objective, result and indicator hierarchy, with 74 standard indicators across the four thematic areas and four pillars; one disaggregation standard by sex, age and disability; and household conversions based on each location’s real household size instead of a fixed multiplier.
- 03
Digital data collection
13 standard KoboToolbox tools widely used across the humanitarian sector, including the Food Consumption Score, reduced Coping Strategies Index, Household Dietary Diversity Score, Livelihood Coping Strategies and RIMA resilience tools. Each is one common tool with skip logic for partner-specific variations, so partners can create more forms as they need them, with offline collection and access set per partner and per country.
- 04
Central MEAL database
A PostgreSQL platform where partners submit quarterly results against shared indicators, reviewed and approved in three steps (submitter, reviewer, approver), with each period locked until the previous one is approved and a full audit trail.
- 05
Live dashboards
Two Power BI dashboards, Reach and Output, and Outcome, connected live to the database, so approved results appear immediately and can be filtered by country, partner, pillar and thematic area.
- 06
Training and handover
A self-paced learning platform with two courses, four live hands-on trainings (KoboToolbox, the MEAL platform, Power BI and Canva) and a structured handover to the donor’s own IT team.
Decisions that made it work
Agree the rules before building
Every structural question was settled in the decision register first, so the forms, database and dashboards were built against one agreed source of truth instead of partner-by-partner interpretation.
Built to grow
Thematic areas and objectives are stored separately even though they match one-to-one today, so new objectives, partners or frameworks can be added without rebuilding the database.
Count people once
Disability is recorded as part of the sex and age totals, never added on top, and each thematic area target is held once rather than repeated under every pillar, which removes the double counting that had crept into consolidated reports.
Fair regional averages
Outcome percentages are combined across projects as weighted averages, so a small project cannot pull the regional figure as much as a large one.
The result
Built with: KoboToolbox, PostgreSQL, Express and Drizzle API, Next.js, Power BI.
Live in all eight countries
Partners report quarterly against one shared indicator set on a single platform.
From Excel to one source of truth
Manual consolidation of partner spreadsheets has been replaced by one approved, auditable database.
Real-time reporting
Approved results flow straight into the donor’s dashboards, ready for management and donor reporting.
Owned by the client
The platform, forms and dashboards run on the donor’s own infrastructure and accounts, with its staff trained to administer them.
Many partners, one view of your data
If your programme runs on spreadsheets from many partners, we can show you what one standard platform would look like in a 30-minute briefing.