Table of Contents
A practical indicator redesign guide using Mosaic as a case study (powered by Wellbi)
Why OVC Program Evaluation Matters (and Why It’s Hard)
In OVC care work, outcomes are complex, long-term, and deeply human. But boards, donors, and management teams still need clear answers: Are our programs working? Where are we drifting? What should we change next quarter?
Most NGOs don’t struggle because they lack activity. They struggle because measurement becomes either too complicated to sustain or too shallow to be useful. This challenge is rooted in real OVC-care organisation experience that we’ve witnessed across countless non-profits. Over the last 12 months at Mosaic, we rebuilt our approach so that evaluation drives decisions!
Mosaic as a Program Evaluation Benchmark
Program focus: family strengthening and aftercare programs.
Scale: ±450 beneficiaries
Data capture: 15–20 staff capture data in the flow of work (often on Wellbi’s mobile phone app)
Analysis capacity: one person (our COO) consolidates and analyses data for quarterly learning and board reporting
Learning rhythm: weekly operational check-ins, quarterly learning reviews, and quarterly board reporting
The key takeaway: you don’t need a big M&E department. You need a disciplined indicator framework, consistent capture, and a repeatable review rhythm.
The Core Shift: Five Organisational Success Indicators
We started with our Theory of Change and forced ourselves to choose five organisational success indicators. Not fifty. Five.
Each key indicator has a small set of sub-indicators underneath it. The key indicator tells you whether you’re winning. The sub-indicators tell you why—and what to fix.
A Practical How-To: Redesign Your Program Evaluation in 7 Steps
1) Lock in the Purpose and Cadence
Before you touch indicators, decide what “success” must enable inside your organisation.
At Mosaic, our purpose was twofold:
Program improvement: identify what needs to change in our delivery (not just whether we’re busy).
Board accountability: build a dashboard our board can use to hold management accountable.
Your output for step 1: one paragraph stating who will use the data, how often, and what decisions it will shape.
2) Use Your Theory of Change to Choose Five Success Indicators
This step is where most NGOs go wrong: they start with donor templates, not their Theory of Change.
We used our Theory of Change to reduce complexity and land on five key indicators for success, each with sub-indicators underneath it. The point is not to measure everything—it’s to measure what actually indicates progress.
Mosaic’s five key indicators, with sub-indicators & cadence (all of these are measurable in Wellbi, with some requiring simple configuration decisions like note templates, surveys, or standardised fields):
Key indicator | Sub-indicator |
Academic progress | Grades vs school average % improving in reading literacy (A–Z Reading) % improving in maths (Numbersense) Grade progression |
Attendance and retention | Aftercare attendance % retained in the programme |
Social and emotional wellbeing | Average SEL survey score % of learners scoring above threshold Behavioural incident counts |
Caretaker capacity and family stability | % of caretakers with improved evaluation score % of participating caretakers with improved score Neglect/abuse referrals Workshop attendance of invited families |
Long-term transitions (EET: education, employment, training) | % youth in EET Caretaker employment / work-study-training-retirement |
Your output for step 2: a one-page table with five key indicators, with sub-indicators, a frequency, and an internal owner.
3) Separate Leading and Lagging Indicators
Leading indicators give early warning (so you can act this month). Lagging indicators confirm results later (so you can prove impact).
Example: attendance and caregiver engagement are often leading indicators; academic progress and EET outcomes are often lagging indicators.
Your output for step 3: Label each sub-indicator as leading or lagging, then decide which ones you review monthly/weekly vs quarterly/annually. You don’t need perfection here; you need a workable distinction so the team knows what to watch weekly vs what to review quarterly.
4) Write a Data Dictionary for Consistent Capture
This is where we hit our biggest constraint.
When we started exporting and analysing our data from Wellbi, we found:
- many missing fields
- qualitative notes that were too cryptic to learn from
This is not a staff-motivation problem. It’s usually a definition + training problem. A data dictionary is a plain-language guide that defines each key field: what it means, how to capture it, and examples of ‘useful’ vs ‘not useful’ entries.
Your output for step 4: a short data dictionary for key indicators plus any narrative fields you rely on for learning.
5) Design for Distributed Data Capture
The breakthrough for us was shifting away from a single M&E person capturing everything. Instead, many team members capture their portion of the work in real time.
Done right, this reduces backlog, improves accuracy, and spreads ownership across the team.
6) Treat Training as an Operating System
Our biggest improvement came from better training. If I could redo the last 12 months, I would:
- train more
- emphasize why the data matters
- show the team how we use it to make decisions
Because when staff understand that the data shapes program decisions (and board reporting), they capture it with more care.
A practical training rhythm that works:
- onboarding training (role-based)
- supervisor refreshers every quarter
- short “data quality moments” in weekly team leader check-ins
- examples of good vs poor capture shown openly (without shaming)
7) Close the Loop with a Repeatable Review Rhythm
Data only matters if it lands in real meetings where decisions happen.
At Mosaic, operations uses dashboards weekly with team leaders (attendance, flags, delivery issues). Donor reporting pulls reliable information without constantly chasing staff. Quarterly, we analyse trends to guide program changes and report to the board.
Your output for step 7: define three recurring moments—weekly operations review, quarterly learning review, and quarterly board dashboard.
What Wellbi Changed for Program Evaluation
Two capabilities made this system practical at Mosaic:
Mobile-friendly distributed capture: staff capture data where the work happens, not days later at a desk.
Flexible exports: we download structured data into Excel quickly for deeper analysis and board-ready reporting.
Wellbi doesn’t replace programme thinking or leadership discipline. It makes a disciplined evaluation system executable with real teams and real workloads. The platform’s comprehensive beneficiary management system enables the kind of real-time data capture that makes sophisticated M&E accessible to smaller organisations.
For organisations implementing this approach, Wellbi provides training and onboarding support to ensure teams can effectively use the evaluation framework from day one.
Two Examples of Data Changing OVC Program Decisions
Engagement decisions: We used Wellbi to visually compare attendance trends, which highlighted patterns of irregular attendance. This led us to introduce a clear attendance protocol and use the system for direct communication with caretakers. By looking at our survey results, we found that checking in with families more often (small, regular support) worked better than doing one big, once-off intervention. It helped improve behaviour, keep families more stable, and keep people involved.
A safety and belonging decision: we introduced a simple child survey in our aftercare programme focused on social and emotional wellbeing—especially whether children feel safe, heard, and respected. We now use that feedback to target staff development (mostly soft skills) and create a “trauma safe” environment at our community centres.
This is what “learning” looks like when evaluation is working: it changes what you do, and how you do it.
Common Program Evaluation Failure Points
Too many indicators: start with five organisational indicators and expand only when capture is stable.
Undefined fields: build the data dictionary and train from it.
Missing fields: make minimum data sets non-negotiable. Supervisors must own quality.
Cryptic notes: standardise templates and examples so narrative data is interpretable.
No decision loop: attach every dashboard to a meeting and an owner.
Building Sustainable OVC Program Evaluation
No NGO is too small for meaningful M&E. In OVC care, we owe it to those we serve to ensure our programs are not simply busy, but genuinely and measurably changing their lives.
If you redesign your indicators from your Theory of Change, define the data clearly, distribute capture across your team, and commit to a review rhythm, you can build an evaluation system that improves programs and strengthens accountability.
For organisations ready to transform their program evaluation approach, implementing this evaluation framework with Wellbi provides the technical foundation to make sophisticated monitoring accessible and sustainable, regardless of team size.
