Ask three people in an NGO what MEAL means and you will get three answers: a reporting obligation for the donor, a job title in the organisation chart, or a set of spreadsheets that someone updates before the quarterly review. All three are partly true, and none of them is the point. MEAL — Monitoring, Evaluation, Accountability and Learning — is the system an organisation builds to know whether its projects are doing what they promised, to let affected people say so, and to change course when the evidence calls for it.

This guide answers the question "what is MEAL?" from the ground up: what each of the four letters produces, how MEAL differs from classic M&E, where it sits in the project cycle and the logframe, what a working MEAL system is made of, who runs it, what donors and auditors expect, where systems usually break, and which tools hold them together.

What is MEAL? Definition and how to set up a MEAL system in an NGO

Reading time: ~17 min

1. MEAL definition: the four letters and what each produces
2. MEAL vs M&E: what the A and the L add
3. Where MEAL sits in the project cycle and the logframe
4. The components of a MEAL system
5. Who does MEAL: roles and responsibilities
6. What donors and auditors expect from MEAL
7. Common MEAL failures
8. MEAL tools: from Excel to MEAL software
9. Mini FAQ
10. Summary

1. MEAL definition: the four letters and what each produces

MEAL stands for Monitoring, Evaluation, Accountability and Learning. It is the set of processes, roles and tools through which an NGO collects evidence about a project, judges that evidence, answers to the people it serves and to its funders, and feeds what it learns back into decisions. The acronym became standard in humanitarian and development organisations during the 2010s, pushed by the Core Humanitarian Standard (CHS) and by donors asking for more than output counts. Each letter has its own logic and its own deliverables.

Monitoring

Monitoring is the routine, continuous collection of data on what the project is doing and delivering: activities completed, outputs produced, indicator values measured, budget consumed. It answers "are we on track?" and it happens throughout implementation, on a fixed calendar. What it produces: dated indicator values against baseline and target, activity progress against the workplan, and monthly or quarterly monitoring reports that let a project manager act before a deviation becomes a problem.

Evaluation

Evaluation is the periodic, structured assessment of a project against explicit criteria, usually the OECD-DAC ones: relevance, coherence, effectiveness, efficiency, impact and sustainability. It answers "did it work, and why?" and it happens at defined moments — baseline, mid-term, final, sometimes ex-post. What it produces: evaluation reports with findings, conclusions and recommendations, and a management response stating what the organisation will do with them.

Accountability

Accountability is the commitment to be answerable to the people the project is meant to serve, and secondarily to partners and donors. In practice it means sharing information about the project with communities, consulting them on design and delivery, and running a complaint and feedback mechanism they can actually use and trust. What it produces: a complaints register with categories, response times and closure status, records of community consultations, and evidence that feedback changed something.

Learning

Learning is the deliberate use of monitoring, evaluation and accountability data to improve the current project and the next one. It answers "what do we now know that we did not know before, and what will we do differently?" What it produces: documented lessons, after-action reviews, learning workshops with a decision log, and visible changes to the next proposal, workplan or standard operating procedure.

The complete MEAL guide goes deeper into each pillar; this article focuses on how the four fit together into one system.

2. MEAL vs M&E: what the A and the L add

For twenty years the sector talked about M&E, monitoring and evaluation, and many organisations still do. French-speaking NGOs say S&E (suivi-évaluation) or, increasingly, SERA — Suivi, Évaluation, Redevabilité, Apprentissage — the French equivalent of MEAL. The vocabulary varies; the substance of the two added letters does not.

M&E measured and judged. It was built around the funder's question: did the project deliver what the contract said? That produced indicator tables and evaluation reports, and it often stopped there. Two things were missing.

• Accountability adds a second audience and a second direction of information. Data no longer flows only from the field up to headquarters and the donor; it also flows from communities into the project, through consultations and complaints, and back out to them as feedback on what was done. The CHS turned this into a verifiable commitment rather than a good intention.
• Learning adds a use for the data beyond reporting. An M&E system can be fully compliant and change nothing. A MEAL system is judged on whether evidence altered decisions: a workplan revised after a monitoring visit, a targeting method changed after complaints, a proposal rewritten after a final evaluation.

In short, M&E is a subset of MEAL. If your organisation already runs solid monitoring and commissions evaluations, moving to MEAL mostly means building the complaint mechanism and the learning loop, and making sure both leave traces.

3. Where MEAL sits in the project cycle and the logframe

MEAL is not a phase; it runs through every phase of project cycle management. During identification and design, MEAL defines the results chain, chooses indicators, estimates baseline needs and budgets the data collection. During implementation, it runs monitoring, handles complaints and produces progress reports. At mid-term it feeds a review that may trigger a workplan or budget amendment. At closure it delivers the final evaluation and the lessons that go into the next design. Treating MEAL as something that starts once activities are underway is the most common design error, and it is usually irreversible: a baseline not collected before the first distribution cannot be reconstructed afterwards.

The logical framework is where MEAL is anchored. Each level of the logframe — overall objective, specific objective or outcome, outputs, activities — carries objectively verifiable indicators, sources of verification and assumptions. Monitoring tracks the output and activity indicators on a short cycle; evaluation examines the outcome and impact levels; the sources of verification column is, in effect, the first draft of the data collection plan. When the logframe and the MEAL plan are built together, every indicator has a method, a frequency and an owner. When they are built apart, the project ends up with indicators that nobody measures and measurements that correspond to no indicator.

4. The components of a MEAL system

A MEAL system is a set of linked components rather than a document. Six of them are necessary for the system to function; everything else is refinement.

MEAL plan and indicator matrix

The MEAL plan turns the logframe into an operational document. For each indicator it specifies the definition and unit, the disaggregation (sex, age, disability, location), the baseline and target, the data source, the collection method and tool, the frequency, the person responsible and the way the value will be used. It also schedules evaluations and reviews and describes the complaint mechanism. The MEAL plan template and method article walks through building one line by line.

Data collection

Monitoring data comes from activity records, distribution lists, attendance sheets, surveys, focus groups, observation checklists and secondary sources. The practical constraints are the ones that decide quality: whether the tool works offline, whether enumerators can be trained on it in a day, whether data entered in the field lands in the same place as the indicator it serves. The offline data collection guide covers the field side.

Data quality

A value is only usable if someone can say where it came from, who collected it, when, and whether it was checked. Minimum data quality practice: a written definition for each indicator, validation rules at entry, spot checks on a sample of records against source documents, and a dated audit trail from the reported figure back to the evidence. Donors increasingly run data quality assessments on exactly this chain.

Complaint and feedback mechanism

The accountability pillar needs at least one channel through which affected people can raise complaints, including sensitive ones (fraud, abuse, exploitation), a register, categories, defined response times, a referral path for safeguarding cases and a closure loop that tells the complainant what happened. The complaint mechanism guide describes the design choices; the AAP and CHS framework article explains the commitments it responds to.

Evaluations against OECD-DAC criteria

Evaluations are planned in the MEAL plan, budgeted, and framed by terms of reference that name the OECD-DAC criteria to be assessed and the questions under each. Baseline, mid-term and final evaluations are the usual sequence; an impact evaluation with a comparison group is the exception, reserved for projects that can afford the design. The impact evaluation article shows what proving results to donors realistically requires.

Learning loops

A learning loop is a scheduled moment where evidence meets decision: a monthly data review where the project manager and the MEAL officer look at indicator trends and complaint categories and decide on actions; a quarterly reflection with field teams; a management response to each evaluation with owners and deadlines; a lessons log consulted when the next proposal is written. Without a calendar and a decision record, learning stays an aspiration.

5. Who does MEAL: roles and responsibilities

MEAL is a shared function with one coordinator. The MEAL officer (chargé·e SERA in French organisations) designs the MEAL plan, builds the tools, trains enumerators, checks data quality, maintains the indicator sheets and the complaints register, drafts the MEAL sections of reports and facilitates learning reviews. In small NGOs the role is often part of the project manager's job or shared across a country office; in larger ones it is a team with a coordinator at headquarters.

• The project manager owns the results. They use monitoring data to steer the project, decide on corrective actions, sign off reported values and carry the management response to evaluations. A project manager who treats MEAL as someone else's paperwork gets reports that describe the project instead of informing it.
• Field teams generate most of the data. Facilitators, distribution teams, community mobilisers and partner staff fill attendance sheets, run surveys, receive complaints in person and observe what indicators do not capture. Their training and their incentive to record accurately are the foundation of data quality.
• Finance is a MEAL partner more often than organisations realise: the cost per output, the budget consumed against indicator progress and the eligibility of MEAL expenses all sit at the junction of the two functions.
• Senior management decides what happens to findings. If evaluation recommendations never reach a management meeting, the learning pillar does not exist, whatever the plan says.

6. What donors and auditors expect from MEAL

Donor expectations converge more than their templates suggest. All of them want a logframe with baselines and targets, a MEAL plan or its equivalent in the proposal, dated indicator values in each progress report, evidence of an accountability mechanism and an evaluation at the end. The differences are in format and emphasis.

• EU and ECHO: the logframe is contractual; ECHO's Single Form asks for indicators with baseline and target per result, a description of the complaint mechanism and of how the organisation meets its AAP commitments; the final report must explain every deviation from target.
• AFD: a logical framework with indicators per objective, a mid-term and final evaluation for larger grants, and growing attention to learning and capitalisation products that outlive the project.
• UN agencies: indicator frameworks aligned with their own country programme indicators, standard reporting cycles, verification visits and, for some agencies, specific quality assurance on partner data.
• Foundations: lighter templates but a strong interest in outcomes and in what the organisation learned, often with more freedom to report in narrative form.

Auditors look at MEAL differently from programme officers. They do not assess whether the project was good; they check whether the figures in the report can be traced to evidence. For an indicator value they will ask for the source document, the date, the person who collected it and any checks performed. For an activity they will cross-check attendance sheets against the figure reported and against the costs charged. For the complaint mechanism they will want the register, not the policy. The donor reporting guide sets out the reporting chain; MEAL is what fills the results half of it.

7. Common MEAL failures

MEAL systems rarely fail loudly. They erode, and the erosion shows up at the final report or the audit. The patterns are familiar to anyone who has inherited a project halfway through.

• Ghost baselines: the proposal gives a baseline value that nobody collected, or that was copied from a secondary source about a different population. Every subsequent "progress" is measured against a number with no origin.
• Orphan indicators: indicators that appear in the logframe and in no data collection tool, or data collected diligently for an indicator that was dropped at the last budget revision. Both come from building the logframe and the MEAL plan separately.
• Parallel spreadsheets: the MEAL officer's tracker, the project manager's dashboard and the finance team's activity list each hold a different version of the same output count. The report takes whichever was open at the time.
• The complaints box nobody opens: a mechanism that exists on paper and in a photo, with no register, no response times and no record of a single complaint being resolved. Auditors and community members both notice.
• Lessons never read: evaluation recommendations filed in a shared drive, a lessons-learned document written at closure and never opened at the next design workshop.
• Values without evidence: a figure updated in the indicator table without the document, the date or the name of the person behind it, which turns the next audit into a reconstruction exercise.

Most of these share one root cause: the components of the system live in different files, maintained by different people, with no structural link between an indicator, its evidence, its budget line and the decision it should inform.

8. MEAL tools: from Excel to MEAL software

Excel is where every MEAL system starts, and it is adequate for a single project with a dozen indicators and one person maintaining the file. It stops being adequate when several projects share indicators, when field data arrives from several collection tools, when values need an attached proof, or when a donor asks who changed a figure and when. Dedicated data collection tools solve the field end but not the tracking end; general-purpose project tools track tasks, not indicators against baselines. The MEAL software comparison sets out the options and what distinguishes them.

How the Abvius MEAL module structures the system

The Abvius MEAL module is built around the components described above, inside the same ERP that holds the project's budget, commitments and documents.

• Logframe: objectives, outcomes, outputs and activities are entered as a tree, and indicators are attached to the level they measure, so the structure of the project and the structure of the MEAL plan are one and the same.
• Indicator sheets: each indicator carries its definition, unit, disaggregation, baseline, target and a series of dated measured values, each with its source, author and attached evidence. Progress against target is computed, not retyped, which removes the parallel spreadsheet problem at the root (see indicator tracking).
• Monitoring calendar: collection events, evaluations and reviews are scheduled per indicator, with reminders to the person responsible, so the MEAL plan has a working calendar rather than a frequency column.
• Complaint form: the complaint form feeds a register with categories, sensitivity flags, response deadlines and closure status, linked to the project it concerns, with access restricted for sensitive cases in line with GDPR and safeguarding rules.
• Budget-indicator link: because budget lines and indicators live in the same system, the cost of an output and the spending rate against indicator progress are available without an export, which is what both the project manager and the donor's efficiency question need.
• Reports: logframe progress tables, indicator histories and complaint statistics are produced per project, per donor and per period, in French or English, with the evidence one click away for the auditor.
• Offline and multi-country: values can be entered in the field without connectivity and synchronised later, and the same structure applies across country offices, which is what a headquarters MEAL coordinator needs in order to consolidate.

Mini FAQ

Is MEAL only for big NGOs?

No. The scale changes, the logic does not. A small association running one project needs a handful of well-defined indicators with baselines, a simple collection routine, one channel for complaints with a register, and a quarterly moment to look at the data and decide. That is a complete MEAL system. What large organisations add is specialised staff, more rigorous evaluations and consolidation across projects.

MEAL vs SERA vs M&E: what is the difference?

SERA is the French acronym for MEAL (Suivi, Évaluation, Redevabilité, Apprentissage); the two describe the same thing. M&E (or S&E in French) is the older, narrower framing that covers monitoring and evaluation only. Many organisations use the terms interchangeably in daily practice, but a donor asking for a "MEAL plan" expects to see accountability and learning addressed explicitly.

How much budget should go to MEAL?

There is no universal percentage, and donors who quote one treat it as a range rather than a rule. The right figure depends on the number and complexity of indicators, the cost of reaching the population (remote areas, security constraints), the evaluations planned and whether an external evaluator is required. The usable method is to cost the MEAL plan line by line — staff time, enumerators, tools, travel, evaluations — and put that in the proposal budget, rather than applying a ratio and hoping it covers the work.

Which indicators should we start with?

Start with the indicators the logframe commits you to, and make sure each has a written definition, a baseline, a target, a method and an owner. Then add one or two outcome indicators that tell you whether the outputs are producing the intended change, and at least one accountability indicator (complaints received and resolved within the deadline). Resist the temptation to track everything that can be counted; ten indicators measured properly are worth more than forty measured occasionally.

Summary

MEAL is the system through which an NGO knows what its projects deliver, judges whether they work, answers to the people they serve and learns from the result. Monitoring produces dated indicator values; evaluation produces judgements against the OECD-DAC criteria; accountability produces a complaints register and consultation records; learning produces decisions that change the next step. The system is anchored in the logframe, runs through the whole project cycle and is made of a MEAL plan, collection tools, data quality checks, a complaint mechanism, planned evaluations and scheduled learning loops, coordinated by a MEAL officer and owned by the project manager. Donors expect to see all of it; auditors expect to trace every reported figure back to evidence. The systems that fail are the ones whose components live in separate files. Whether you call it MEAL, SERA or still M&E, the test of what MEAL is in your organisation is simple: can you show where each figure comes from, what people told you, and what you changed because of it? If the honest answer is spreadsheets and goodwill, start a free trial of Abvius and put the logframe, the indicators, the evidence and the complaints in one place.