Useful evaluations answer real management questions
An endline evaluation should not simply repeat the project logframe. It should answer the questions that matter to programme teams, donors, partners and communities. These questions often include what changed, for whom, why, at what cost, under which conditions and with what implications for future action.
Results need explanation, not only measurement
Indicator tables are important, but they are rarely sufficient. A useful endline evaluation explains why some targets were achieved and others were not. This requires triangulation of survey data, qualitative evidence, implementation records and stakeholder perspectives.
Recommendations should be specific and responsible
Many evaluation reports contain recommendations that are too broad to implement. A stronger approach is to identify priority actions, responsible actors, timing, feasibility and expected contribution to future results. This makes the report more useful for management response and programme design.
Limitations should be transparent
Every evaluation has limitations. These may relate to sampling, recall, missing baseline data, access constraints, attribution, comparison groups or data quality. Credible reporting acknowledges these limitations and explains how they affect interpretation.
Learning products increase use
A final report is important, but clients often need additional products such as presentations, executive summaries, learning briefs and recommendation matrices. These products make evidence easier to discuss and apply.
Compare results with the right reference points
An endline should examine change against the baseline, targets and relevant implementation evidence rather than reporting endline percentages in isolation. Where measurement changed between rounds, the report should explain whether direct comparison remains valid and what limitations apply.
Useful interpretation distinguishes statistical change, programme achievement and participant perception. These are related but not identical forms of evidence, and combining them carelessly can overstate conclusions.
Explain why results changed
Clients need more than a judgement that an indicator was achieved or missed. The evaluation should examine implementation quality, contextual shocks, targeting, service uptake, partnerships and other mechanisms that help explain the observed pattern.
Qualitative evidence and programme records are particularly valuable here. They can show why the same intervention worked differently across locations or participant groups and identify practical lessons for future design.
Use proportionate causal language
Most NGO endline evaluations cannot prove that the programme alone caused every observed outcome. The strength of causal claims should reflect the evaluation design, availability of comparison evidence and the credibility of alternative explanations.
Contribution analysis, triangulation, process evidence and comparison with the programme theory of change can support strong conclusions without pretending that a non-experimental design provides experimental certainty.
Make management response possible
Recommendations should identify what should change, who is responsible and which actions deserve priority. Where recommendations imply additional resources, policy change or further analysis, the report should say so rather than presenting them as simple operational fixes.
A recommendation matrix or management-response process helps the client track which actions are accepted, adapted or deferred. This turns the endline from a compliance product into an organisational learning tool.