Introduction
In the non-profit sector, it is no longer enough for an organisation to say:
We implemented 20 activities, distributed 5,000 packages, and trained 300 beneficiaries.
These figures are important, but they tell us what we did, not necessarily what changed as a result of what we did.
This is the difference between an organisation that collects data to prepare reports and one that uses data to understand its programmes, make better decisions, and demonstrate results through evidence that can be verified.
This is where Monitoring, Evaluation, Accountability and Learning (MEAL) becomes essential.
MEAL is not simply a collection of forms. It is not a department responsible for gathering figures at the end of each month, nor is it merely a reporting system.
MEAL is a system for producing evidence, and reporting is one of its outputs.
When properly designed, it connects:
What we want to change → What we will do → How we will measure it → What data we will collect → What the results show → What we learned → What we will change
A strong MEAL system enables an organisation to monitor implementation, measure results, listen to beneficiaries, verify data quality, interpret variances, capture lessons, and turn knowledge into decisions.
It also gives donors something more important than a well-designed report:
Evidence that can be traced and verified.
This article builds on the original framework covering the components of MEAL, results frameworks, indicators, data quality, data collection tools, data management, reporting, organisational learning, and the relationship between MEAL, governance, risk management, and sustainability.
First: What Is a MEAL System?
MEAL brings together four interconnected components:
Monitoring
The regular collection of data to determine whether implementation is progressing according to plan and whether indicators are moving in the expected direction.
Evaluation
Assessing the extent to which a project has achieved its intended results and objectives, understanding the reasons for success or underperformance, and identifying lessons.
Accountability
Ensuring that beneficiaries and stakeholders have access to appropriate information, opportunities to participate, and safe channels for feedback and complaints, while ensuring that what the organisation learns from them is used to improve its work.
Learning
Turning data, evaluations, and experience into knowledge that informs future programmes, policies, and decisions.
These components do not operate separately:
Monitoring tells us what is happening.
Evaluation helps us understand why it happened.
Accountability brings in the voices of affected people.
Learning turns what we know into improvement.
Second: Why Do Donors Distrust Some Reports?
The problem is often not the appearance of the report, but what lies behind the numbers it contains.
A report may contain dozens of pages, photographs, and tables and still be weak if it cannot answer basic questions:
What was achieved?
Compared with what?
How was it measured?
Where did the number come from?
Can it be verified?
Why were some targets not achieved?
What did you learn?
What will you change?
Common weaknesses include:
- Focusing on activities rather than results.
- Lack of a clear baseline.
- Poorly defined indicators.
- Incomplete or undocumented data.
- Presenting numbers without interpreting them.
- Hiding variances and challenges.
- Failing to identify data sources.
- Collecting feedback without showing how it was used.
- Overstating the extent to which observed changes can be attributed to the project.
A credible report is therefore not one in which every indicator looks successful.
A credible report is one that can substantiate its results and objectively explain both successes and shortcomings.
Third: Start With the Logic of Change Before Collecting Data
A common mistake is to begin designing surveys and indicators before clearly agreeing on what change the project is trying to achieve and how that change is expected to happen.
Three interconnected levels can help:
1. Theory of Change
Explains why and how the intervention is expected to lead to the desired change, including the assumptions underlying that logic.
2. Results Framework / Logframe
Translates that logic into interconnected results that can be managed and monitored.
3. MEAL Plan
Defines how the organisation will measure those results:
What will we measure? Where will the data come from? How will it be collected? When? And who is responsible?
The sequence therefore becomes:
Theory of Change → Results Framework / Logframe → Indicators → MEAL Plan → Data → Analysis → Decisions & Reporting
If the project logic itself is unclear, collecting more data will not solve the problem.
Fourth: The Results Framework — Do Not Confuse What You Did With What Changed
A simplified results chain can be represented as:
Inputs → Activities → Outputs → Outcomes → Impact
Inputs
The human, financial, and technical resources used.
Activities
The actions implemented by the project.
Outputs
The immediate deliverables resulting from those activities.
Outcomes
Changes in knowledge, behaviour, access to services, or circumstances resulting from the intervention.
Impact
The broader or longer-term change to which the project contributes.
Suppose a project provides vocational training to 200 young people.
Activity: Deliver vocational training.
Output: 180 participants completed the training.
That alone is not sufficient to claim that the project achieved impact.
We can move further to:
Outcome: Percentage of participants who acquired the targeted skills.
Then:
Outcome: Percentage of participants who obtained employment or generated a source of income within six months.
This is where the real transition begins:
From:
What did we deliver?
To:
What changed?
Fifth: Monitoring and Evaluation — Two Different Questions
Monitoring and evaluation complement each other, but they are not the same.
Monitoring
Monitoring is continuous throughout implementation.
It helps the organisation determine:
- Were activities implemented?
- Are we reaching the intended groups?
- Are we on schedule?
- Are indicators moving as expected?
- Is there a variance requiring intervention?
Evaluation
Evaluation provides deeper analysis and asks:
- Did the project achieve its intended results?
- Why did it succeed or fail?
- What made the difference?
- What lessons were learned?
- Are the results likely to be sustained?
An evaluation may take place during implementation, at project completion, or after a period of time, depending on its purpose.
Monitoring helps us manage the project while it is running; evaluation helps us understand its value and results more deeply.
Sixth: An Indicator Alone Is Not Enough
A strong MEAL system does not stop at naming an indicator.
For example:
Percentage of beneficiaries whose knowledge improved.
This is insufficient if we do not define what “improved” means.
A stronger formulation would be:
Percentage of participants who achieved at least a 20% improvement between pre-test and post-test scores.
Then we establish:
Baseline: 35%
Target: 70%
Actual: 64%
Now we have something that can be measured, compared, and analysed.
An indicator therefore generally requires:
Indicator → Definition → Baseline → Target → Data Source → Collection Method → Frequency → Disaggregation → Responsibility
The key principle is:
Do not measure only what is easy to measure; measure what you need to know to make decisions and demonstrate results.
Seventh: Indicator Reference Sheets — Make Sure Everyone Measures the Same Thing
As the number of projects and teams increases, so does the risk that different people will interpret the same indicator differently.
This is where the Indicator Reference Sheet (IRS) becomes valuable.
For each indicator, it may define:
- Indicator name.
- Definition.
- Purpose.
- Calculation method.
- Numerator and denominator for percentage indicators.
- Unit of measurement.
- Data source.
- Data collection tool.
- Measurement frequency.
- Person responsible for collecting and reviewing data.
- Required disaggregation.
- Methodological limitations or notes.
For example, if the indicator is:
Percentage of participants who completed the training.
Everyone should understand:
What does “completed” mean?
Is attending 60% of sessions sufficient? 80%? Or must participants also pass a final assessment?
A standardised definition prevents inconsistent results across staff, projects, and reports.
Eighth: Baseline, Target, and Actual — Numbers Need Context
An organisation reporting:
We achieved 68%.
tells us very little on its own.
If:
Baseline = 30%
Target = 60%
Actual = 68%
that tells a very different story from:
Baseline = 65%
Target = 90%
Actual = 68%
Results should therefore always be interpreted within context:
Baseline → Target → Actual → Variance → Explanation
When a variance exists, it should not be hidden.
Instead, ask:
Why did it happen, and does it require corrective action?
Ninth: Data Quality — Can You Prove the Number?
The best analysis in the world cannot rescue poor-quality data.
Organisations therefore need Data Quality Assurance / Data Quality Assessment (DQA) processes covering dimensions such as:
Accuracy — Completeness — Timeliness — Consistency — Integrity
Data should be:
- Accurate.
- Complete.
- Available when needed.
- Consistent across sources and reporting periods.
- Protected against unauthorised alteration or manipulation.
Important practices include:
- Standardised data collection tools.
- Training data collectors.
- Reviewing missing or implausible values.
- Checking for duplicates.
- Periodic database reviews.
- Field verification where appropriate.
- Documenting corrections.
- Retaining supporting evidence.
There is a very simple test:
Select a number from the report. Can you trace it back to the database and then to the original record or collection tool from which it came?
This is Data Traceability.
The clearer the path of a reported figure, the better positioned the organisation is to demonstrate the credibility of its data to donors and auditors.
Tenth: Do Not Look Only at the Average — Who Actually Benefited?
An overall figure can conceal important differences between groups.
Suppose a project reports that 75% of beneficiaries achieved the intended result. Performance may appear excellent.
But what if the result was:
85% in one location compared with 52% in another?
This is where Disaggregation becomes important.
Depending on the programme and the legitimate purpose for collecting the data, results may be analysed using appropriate variables such as:
- Age.
- Gender.
- Location.
- Disability.
- Type of intervention.
- Other categories relevant to programme objectives.
The objective is not to collect as much personal information as possible.
It is to answer:
Who benefited, and who did not benefit to the same extent?
That question can lead to better and more equitable decisions.
Eleventh: Data Collection Tools — Choose the Tool Based on the Question
There is no single tool suitable for every type of information.
Surveys
Useful for structured data and for measuring knowledge, satisfaction, or other quantifiable changes.
Interviews
Useful for understanding experiences, causes, and context.
Focus Group Discussions
Useful for exploring the perceptions of a particular group and discussing issues requiring deeper interpretation.
Field Observation
Useful for verifying implementation and assessing service quality in the actual delivery environment.
Administrative Records
These may include attendance records, service records, and project databases.
Important findings should not always depend on a single source.
If a survey identifies an important result and that result is also supported by interviews, observations, and administrative records, confidence in the interpretation increases.
This is the essence of Triangulation:
Using more than one source or method to verify and better understand a finding.
Twelfth: Protecting Beneficiary Data Is Part of Good MEAL
Collecting more data does not necessarily mean having a better MEAL system.
Every piece of personal information collected creates an additional responsibility to protect it.
Organisations should therefore consider principles such as:
Data Minimisation: Collect only the data that is necessary.
Informed Consent: Explain the purpose of data collection appropriately.
Access Control: Define who can access the information.
Secure Storage: Protect data during storage and use.
Retention: Define how long information will be retained.
Anonymisation / Pseudonymisation: Reduce the link between data and identifiable individuals where appropriate.
Before adding a question to a data collection tool, it is useful to ask:
Why do we need this information, and what will we do with it?
If there is no clear answer, there may be no reason to collect it.
Thirteenth: Data Management — From Scattered Files to a Reliable Source of Information
Data management does not end when information is entered into Excel or another system.
The organisation needs to define:
- Who collects the data?
- Who reviews it?
- Where is it stored?
- Which version is authoritative?
- Who can modify it?
- How is it backed up?
- Where is supporting evidence stored?
- How are corrections documented?
- When is data archived or deleted?
Having a clear Single Source of Truth reduces discrepancies between figures used by programme, finance, MEAL, and management teams.
It also reduces the time wasted at the end of every reporting period trying to answer the familiar question:
Which number is correct?
Fourteenth: Attribution or Contribution? Do Not Claim More Impact Than You Can Demonstrate
This is one of the most important credibility issues in impact reporting.
Suppose household incomes improve after an economic programme.
Does that automatically mean the project caused the improvement?
Not necessarily.
Other factors may have contributed, including market changes, other programmes, or wider economic and social developments.
It is therefore important to distinguish between:
Attribution
Demonstrating, with an appropriate methodology, that the observed change can be attributed to the intervention.
Contribution
Demonstrating that the intervention reasonably and credibly contributed to the observed change while recognising the influence of other factors.
Reporting language should therefore match the strength of the methodology and evidence.
Credibility does not come from claiming the greatest possible impact.
It comes from:
Saying what the evidence can actually support.
Fifteenth: How Does a Number Become a Decision?
The real value of data begins when it influences a decision.
Suppose a project aims for 90% of participants to complete a training programme.
After three months:
Target: 90%
Actual: 72%
Instead of simply recording the figure and waiting for the final report, analysis should begin:
Why?
The data may reveal that most dropouts occur during evening sessions because participants face transportation difficulties.
Now we have a clear chain:
Data → Pattern → Analysis → Decision → Action
The decision may be to:
Change the training schedule or provide an appropriate transport solution.
The organisation can then measure:
Did completion rates improve?
At this point, MEAL becomes a project management tool rather than merely a way of documenting what happened after implementation.
Sixteenth: How Does a Donor Read the Report?
Instead of structuring the report around a list of activities, its core should be organised around results.
For each important indicator or result, a useful logic is:
Target → Actual → Variance → Evidence → Explanation → Corrective Action / Next Step
For example:
| Indicator | Target | Actual | Variance |
|---|---|---|---|
| Training completion rate | 90% | 82% | -8% |
But the table alone is not enough.
The report should explain:
Why did the variance occur?
What evidence supports the explanation?
Did it affect other results?
What did the organisation do?
What will it do during the next reporting period?
This leads to an important principle:
A donor-trusted report is not one that hides the red indicators; it is one that can explain and respond to them.
Seventeenth: What Should a Professional Report Include?
Depending on the project and donor requirements, a report may include:
- Executive summary.
- Objectives and key results.
- Indicators and targets.
- Target-versus-Actual comparisons.
- Explanation of variances.
- Analysis of results rather than descriptions of activities alone.
- Challenges and risks.
- Corrective actions.
- Beneficiary feedback where relevant to results.
- Lessons learned.
- Recommendations and next steps.
- Relevant evidence and annexes.
The language should be:
Accurate — Objective — Verifiable — Free from Exaggeration
A statement such as:
“The project achieved exceptional and unprecedented impact.”
has little value without evidence.
Whereas:
“The indicator increased from 42% at baseline to 67% at endline, against a target of 65%.”
gives the reader something that can be assessed.
Eighteenth: Accountability Adds What Numbers Cannot Tell Us
MEAL is not complete if it only monitors indicators.
An organisation may achieve its targets while beneficiaries still have a poor experience.
Accountability therefore adds another important question:
What do the people affected by the programme say?
Participation, feedback, and complaints channels can reveal issues that may not appear on a dashboard, such as:
- Difficulty accessing services.
- Unclear eligibility criteria.
- Problems with service delivery.
- Needs that were not anticipated during project design.
- Unintended effects.
When this information leads to programme adjustments, accountability becomes part of learning and decision-making, rather than merely a mechanism for receiving complaints.
Nineteenth: Learning — The Element That Makes MEAL Worth the Investment
If a project ends, its report is archived, and nothing changes within the organisation, much of the value of MEAL has been lost.
Organisational learning means systematically asking:
What worked?
What did not work?
Why?
What should we repeat?
What should we stop doing?
What will we do differently?
This can be supported through:
- Periodic review meetings.
- After-Action Reviews.
- Learning Logs.
- Documentation of lessons learned.
- Knowledge sharing between teams.
- Updating procedures.
- Incorporating lessons into the design of future projects.
Most importantly, organisations should move from:
Lesson Identified
to:
Lesson Applied
A lesson that is documented but never influences a decision does not represent complete organisational learning.
Twentieth: How Do You Know Whether a MEAL System Is Mature?
Maturity does not mean having sophisticated software or a large team.
It is demonstrated by the organisation’s ability to connect planning, data, evidence, and decision-making.
Evidence of maturity may include:
- Theory of Change.
- Results Framework / Logframe.
- MEAL Plan.
- Clearly defined and measurable indicators.
- Indicator Reference Sheets.
- Documented baselines and targets.
- Approved data collection tools.
- Structured databases.
- Data Quality procedures.
- DQA records.
- Data protection controls.
- Feedback and complaints data.
- Periodic analytical reports.
- Evaluation reports.
- Learning Logs.
- Corrective Action Tracker.
- Evidence that findings influenced decisions.
The organisation does not need to create documents simply to increase the number of documents it holds.
Their value lies in forming an interconnected evidence chain.
Twenty-First: MEAL, Governance, and Risk Management
MEAL is not isolated from organisational management.
Reliable data helps boards and executive management understand:
Are programmes achieving their intended results?
Where are the variances?
What risks are emerging?
Are resources being directed toward the most effective interventions?
Are there unintended results?
MEAL therefore supports:
Governance by providing reliable information for oversight and decision-making.
Risk Management by enabling early identification of problems and variances.
Accountability by making performance visible to beneficiaries, donors, and other stakeholders.
Sustainability by helping the organisation identify what should be continued, scaled, redesigned, or discontinued.
Twenty-Second: How Can a Donor Know That the MEAL System Actually Works?
An organisation may have an excellent MEAL Policy, but donors need evidence of implementation.
Such evidence may include:
Theory of Change — Logframe — MEAL Plan — Indicator Reference Sheets — Baseline Data — Targets — Data Collection Tools — Databases — DQA Records — Source Documents — Beneficiary Feedback — Evaluation Reports — Learning Logs — Corrective Action Tracker — Donor Reports
But the maturity test can be reduced to one powerful request:
Select a number from your latest report and prove it.
Can the organisation explain:
Where did the number come from?
How was it calculated?
What is the indicator definition?
When was the data collected?
Who collected and reviewed it?
What is the original supporting evidence?
Can the calculation be repeated and produce the same result?
Then comes the second question:
What did you do as a result of what this data told you?
The first tests data reliability.
The second tests the organisational value of MEAL.
Before Moving to the Next Article...
An organisation can begin with practical steps:
✓ Review its Theory of Change and Results Framework.
✓ Ensure a clear distinction between Outputs, Outcomes, and Impact.
✓ Establish Baselines and Targets for key indicators.
✓ Develop Indicator Reference Sheets.
✓ Build a clear MEAL Plan.
✓ Review data collection tools.
✓ Define appropriate levels of Disaggregation.
✓ Establish Data Quality and DQA procedures.
✓ Ensure that reported figures can be traced to their original sources.
✓ Review controls for protecting beneficiary data.
✓ Use Triangulation for important findings where appropriate.
✓ Analyse variances rather than hiding them.
✓ Connect beneficiary feedback to decision-making.
✓ Establish a Learning Log and Corrective Action Tracker.
✓ Structure reports around results and evidence rather than activities alone.
Quick Self-Assessment
Ask yourself:
□ Can we explain how the project is logically expected to produce its intended results?
□ Do we clearly distinguish between Output, Outcome, and Impact?
□ Does every indicator have a standard definition?
□ Do key indicators have Baselines and Targets?
□ Do we know the source of every number reported?
□ Can each reported figure be traced to its original evidence?
□ Do we periodically review data quality?
□ Do we analyse results by relevant groups rather than relying only on totals?
□ Do we use multiple sources to verify important findings where appropriate?
□ Do we collect only the data we need and protect it appropriately?
□ Do we distinguish between demonstrating Attribution and demonstrating Contribution?
□ Do we explain variances objectively?
□ Has data led to an actual change in a programme or decision during the past year?
□ Can we demonstrate that change to a donor?
If the answer is “No” to several of these questions, the organisation may be collecting a great deal of data, but it may not yet have built an integrated MEAL system.
Conclusion
The real value of MEAL is not measured by the number of surveys, indicators, or reports an organisation produces.
It lies in the organisation’s ability to create a reliable chain that begins with the right question and ends with a better decision:
Clear Theory of Change → Defined Results → Measurable Indicators → Reliable Data → Objective Analysis → Learning → Decision → Improvement
A mature organisation does not use data to prove that it was always right.
It uses data to discover where it succeeded, where it fell short, who benefited, who did not, and what needs to change.
This is why a donor-trusted report does not begin when the report is written.
It begins with project design, indicator definition, measurement systems, and the protection of data quality.
When an organisation can substantiate a figure, explain a result, acknowledge a variance, show what it learned, and demonstrate how a decision changed because of the evidence, it moves beyond simply collecting data toward managing knowledge and impact.
At that point, MEAL becomes more than an oversight function or a donor requirement.
It becomes part of how the organisation determines whether its mission is actually being achieved — and how it can achieve it better.
