Monitoring and Evaluation for NGOs: Results Frameworks, Evidence and Practical Evaluation
August 2, 2026 · Youth Development · 6 min read
Monitoring and evaluation are related but different parts of programme management. Monitoring tracks implementation and results over time. Evaluation makes a structured judgement about an intervention, including what changed, for whom, why and whether the intervention contributed to those results.
A useful M&E system starts with clear decisions and learning questions. It should not be justified with unsupported claims such as “47% of NGO programmes have inadequate M&E”, “strong M&E triples donor renewal” or “68% of programmes do not measure long-term impact”.
Monitoring and evaluation are not the same activity
OECD guidance distinguishes ongoing monitoring from periodic evaluation. Monitoring tracks inputs, activities, outputs and other agreed indicators during implementation. Evaluation is a more structured assessment of an intervention’s design, implementation and results.
Source: OECD, Evaluating development co-operation, July 2026
Both are useful, but they answer different questions.
Start with a theory of change
A theory of change explains how programme activities are expected to contribute to outcomes and why those links are plausible. It helps teams distinguish what they deliver from the changes they hope to create.
A simple chain might include:
- Inputs: money, people, equipment and partnerships;
- Activities: what the programme does;
- Outputs: the immediate products or services delivered;
- Outcomes: changes in behaviour, capability, access, practice or systems;
- Impact: broader or longer-term change to which the intervention may contribute.
The World Bank’s current M&E learning resources place theory of change and results frameworks at the centre of results-focused project design.
Source: World Bank, Monitoring, Evaluation and Results
Build a results framework from the decisions you need to make
An indicator should exist because someone will use the information. Before collecting data, ask:
- What decision will this measure inform?
- What result does it represent?
- Who is responsible for collecting it?
- How often is it needed?
- What is the baseline?
- What data-quality risks exist?
- How will disaggregation be handled safely?
A long indicator list can create reporting burden without improving decisions.
Separate outputs from outcomes
Common monitoring data such as people trained, sessions delivered, materials distributed or grants issued are outputs. They show implementation volume.
They do not by themselves show whether capability, behaviour, access, income, wellbeing or institutional performance changed. Programmes should identify outcome measures that match their actual objectives.
Use the OECD DAC evaluation criteria thoughtfully
The OECD DAC defines six evaluation criteria:
- Relevance: is the intervention doing the right things?
- Coherence: how well does it fit with other interventions and context?
- Effectiveness: is it achieving its objectives?
- Efficiency: how well are resources being used?
- Impact: what difference does it make?
- Sustainability: will the benefits last?
Source: OECD DAC Evaluation Criteria
The OECD explicitly warns against applying the criteria mechanically. They should be selected and interpreted according to the evaluation purpose, intervention and stakeholder needs.
Choose methods to match the question
No single M&E method is appropriate for every programme. Depending on the question, a team may use:
- administrative records;
- surveys;
- interviews and focus groups;
- observation;
- case studies;
- routine service data;
- financial or operational data;
- comparison groups or experimental/quasi-experimental approaches where feasible and ethical;
- contribution analysis or theory-based evaluation where causality is complex.
The method should be proportionate to the programme, the decision and the consequences of getting the answer wrong.
Do not claim attribution without an appropriate design
If employment rises after a training programme, the change may also reflect labour-market conditions, participant selection, other services or economic changes. A before-and-after comparison alone does not automatically prove the programme caused the result.
Evaluation should distinguish:
- what changed;
- whether the programme plausibly contributed;
- what other explanations exist;
- how confident the evaluator can be in the conclusion.
Data quality matters
Useful M&E depends on definitions that are consistent and data that can be checked. Programme teams should document:
- indicator definitions;
- data sources;
- collection frequency;
- responsibility;
- missing-data rules;
- quality checks;
- changes to methods over time.
Mobile data-collection tools can improve speed and field access, but technology does not automatically make data accurate.
Protect participants and sensitive data
NGO and development programmes may collect information about health, income, disability, legal status, safeguarding or vulnerable groups. M&E plans should apply appropriate consent, privacy, security and data-minimisation practices.
Collect only what the programme can protect and use responsibly.
Design evaluation for use, not only for donor compliance
The OECD’s 2026 guidance emphasises evaluation as a source of credible evidence for decisions, learning, accountability and stronger development impact. Reports are more likely to be used when evaluation questions are connected to real programme decisions and findings arrive while action is still possible.
Source: OECD, Evaluating development co-operation, 2026
A practical M&E cycle
- Clarify the programme objective and theory of change.
- Define the decisions and learning questions.
- Select a small set of useful indicators.
- Establish baselines and data responsibilities.
- Monitor implementation and outcomes.
- Investigate unexpected results and data-quality problems.
- Plan evaluations around the questions that require deeper evidence.
- Use appropriate methods and document limitations.
- Communicate findings to people who can act on them.
- Feed learning back into programme design.
Good M&E is not the largest dashboard or the longest donor report. It is an evidence system that helps a programme understand whether it is doing the right work, whether results are emerging and what should change next.
Turn the M&E framework into a management routine
The framework becomes useful only when someone reviews the evidence and makes a decision. A practical operating rhythm can be built around a small number of recurring questions: What changed since the last review? Which indicators are moving away from target? Is the cause a delivery problem, a weak assumption, a data-quality issue or an external change? What action should follow, who owns it and when will the team know whether it worked?
For each important indicator, define a decision rule before the data arrives. A large drop in attendance might trigger a participant follow-up and accessibility review. A placement indicator may require segmentation by location or participant group before management decides whether the programme model is working. A recurring data-quality problem may require a process fix rather than another training session for field staff.
Build an indicator reference sheet, not just an indicator list
Each core indicator should have an operational definition that different staff can apply consistently. Record the numerator and denominator where relevant, the data source, collection method, frequency, disaggregation, responsible role, quality checks and any known limitations. This prevents teams from discovering halfway through a programme that two offices have been counting the same result differently.
Also define what the indicator cannot tell you. Completion data can show whether participants stayed through an activity, but not whether capability improved. Employment placement can show an immediate transition, but not retention or job quality. Strong M&E combines several signals rather than forcing one measure to carry the whole story.
Use evaluation questions to focus the evidence
Before commissioning an evaluation, write the decisions the evaluation needs to support. Questions may address relevance, implementation quality, reach, effectiveness, equity, sustainability or why outcomes differed across groups. This keeps the evaluation from becoming a large report with no clear management use.
A proportionate evaluation plan matches method to importance and uncertainty. Routine monitoring may answer straightforward delivery questions. Qualitative interviews can explain why an outcome pattern occurred. More rigorous comparison designs are justified when causal attribution matters enough to warrant the additional cost and complexity.
Related MATSH resources
For the programme-design side of the same work, see Youth Development Programs: Principles, Evidence and How to Design Them Well.
Sources
Need In-House Training?
We run all our courses as private programmes for organisations across the GCC and Africa.
Request In-House →