Indicators and the MEAL plan

1What makes a good indicator?

An indicator is a measure that shows whether change is taking place. It must be SMART: specific, measurable, achievable, relevant and time-bound.

Weak: “improved hygiene”. SMART: “% of targeted households in which a member can name at least 3 of the 5 critical times for handwashing, rising from 20% to 60% by December 2026”.

• Quantitative (number, %, average) and qualitative (perceptions, level of satisfaction, measured using scales or interviews).
• Disaggregation: by sex, age and disability as a minimum. For disability, the standard is the Washington Group Short Set of Questions (6 questions on functional difficulties).
• Standard indicators: use those of the cluster, Sphere or the donor wherever possible, so that results can be compared and aggregated.

2Baseline, target and reference sheet

Each indicator has a baseline value (the starting situation, measured by the baseline study) and a target (the value aimed for by a given date). Without a baseline, change cannot be measured.

The indicator reference sheet (PIRS, Performance Indicator Reference Sheet) prevents everyone from measuring in their own way. It specifies: the exact definition and terms, the numerator and denominator, disaggregation, the data collection method and tool, the frequency, the person responsible, the baseline and target, and known limitations.

3The indicator tracking table

The tracking table (indicator tracking table, ITT) lists all the logframe indicators with, for each one, the baseline, the target, the values achieved per period and the percentage of achievement. A simple reading rule: green above 90% of the target expected to date, amber between 70 and 90%, red below. Every significant variance must have an explanation and a corrective action.

4The MEAL plan

The MEAL plan is the user manual for the system. It contains:

1. The logframe and the list of indicators with their reference sheets.
2. The timeline: baseline study, monthly monitoring, post-distribution monitoring, quarterly reviews, endline study, evaluation.
3. The tools: questionnaires, checklists, complaint forms.
4. The roles: who collects, who checks, who analyses, who decides.
5. The data flow: from collection to the database, then to reports and decision-making meetings.
6. The complaints and feedback mechanism.
7. The MEAL budget and human resources.
8. The data protection rules.

5Choosing indicators: standard sources and reference thresholds

The first rule for a MEAL professional is not to invent an indicator when a standard indicator exists. Standard indicators have proven definitions, documented measurement methods and recognised interpretation thresholds; they allow comparison with other projects and feed into cluster dashboards. The main sources are the Sphere Handbook, cluster monitoring frameworks (often included in the humanitarian response plan), the Indicator Registry managed by OCHA, donor key indicators (ECHO, BHA) and standard survey modules (for example those of the World Food Programme for food security, or SMART nutrition surveys).

Commonly used indicators and reference thresholds

Sphere thresholds are key indicators intended to measure whether a standard is being met, not absolute norms. The Handbook asks for them to be adapted to the context and for the gap and its reason to be documented when they cannot be met. Writing in a report "the Sphere standard is not met because... and this is what we are doing" is professional practice; ignoring the gap is not.

Proxy indicators and qualitative indicators

When the intended change cannot be measured directly, a proxy indicator (proxy) is used: the presence of soap and water at the handwashing place (observation) is a more reliable indicator of practice than the respondent's own statement. Qualitative changes (sense of safety, trust in the complaints mechanism, social cohesion) are measured with scales (for example a Likert scale from 1 to 5) with strictly identical wording between baseline and endline, or with structured qualitative methods whose results are coded.

How many indicators?

A 12- to 24-month project rarely exceeds 10 to 15 logframe indicators. Each additional indicator has a cost: collection, verification, analysis, reporting. Beyond that, a distinction is made between logframe indicators (accountability to the donor) and internal management indicators (activity monitoring, process quality), which appear only in the MEAL plan.

6Counting people reached: rules, disaggregation and double counting

"How many people did you reach?" is the most frequently asked and the most poorly measured question. People-reached figures feed cluster dashboards and the monitoring of the humanitarian response plan; repeated errors distort the coverage analysis of the entire response.

Definitions to set before any counting

• Person reached: a person who actually received at least one good or service from the project, according to the activity definition (a household receiving a kit counts for its actual size, established at registration, not for a theoretical average size).
• Unique people: each person is counted only once, regardless of the number of services received. Unique people are calculated from a database with identifiers, never by adding up the figures for each activity.
• Direct and indirect beneficiaries: indirect beneficiaries (the population of a village served by a rehabilitated water point) must be estimated using a written method and presented separately.
• Reached and targeted: the cluster and the donor compare the people reached with the people targeted by the project; any significant gap, above or below, must be explained.

Example of calculating unique people

By cross-checking household identifiers in the database, it turns out that 950 households received both cash transfers and kits, and that 3,600 of the session participants belong to households already counted. The number of unique people is therefore: 7,320 + 8,700 − (950 × average size of the households concerned, here 6.0, i.e. 5,700) + (4,100 − 3,600) = 10,820 unique people. Reporting 20,120 people would be an 86% overestimate. Auditors and cluster coordinators know this error and ask for the calculation method.

Disaggregation in practice

The minimum disaggregation expected by donors and clusters is by sex and age group, often 0-5 years, 6-17 years, 18-59 years and 60 years and over (the exact age brackets depend on the cluster and the donor: check them before designing the registers). Disability is increasingly added as a matter of course, measured with the Washington Group questions, in line with the IASC Guidelines on Inclusion of Persons with Disabilities in Humanitarian Action (2019).

• When only household size is known, the breakdown by sex and age is estimated from the baseline proportions or official demographic data; the estimation method must be written in the reference sheet.
• The Washington Group questions ask about functional difficulties (seeing, hearing, walking, remembering or concentrating, self-care, communicating), never "are you disabled?", which strongly underestimates prevalence.
• The Short Set is validated for people aged 5 and over; for children, the Child Functioning Module developed with UNICEF is recommended.
• A reported disability rate far below global estimates (WHO estimates that about 16% of the world's population lives with a significant disability) should raise suspicion of a data collection or access problem.

7Full example: indicator reference sheet and indicator tracking table of a project

Fictional context: a 12-month food security project by the NGO "Terres d'Avenir" in the Moussoro-Nord region, benefiting 1,800 displaced and host households, through monthly cash transfers for 6 months and livelihoods support.

The reference sheet for the main outcome indicator

The indicator tracking table at month 6

The variance analysis accompanying the table

• SO.1 (amber): disaggregation shows 58% among host households, but 44% among displaced households and 39% among households including a person with a disability. Focus group discussions indicate that displaced households spend part of the transfer on rent, which has risen by 30% in six months. Proposed measures: share the analysis with the Cash Working Group to review the transfer value against the minimum expenditure basket, and consider a top-up for households including a person with a disability.
• R1.2 (amber): two cycles were delayed by a liquidity shortage at the payment service provider. Measure: a strengthened payment deadline clause in the contract and a back-up provider identified.
• R2.1 (red): delayed procurement of agricultural inputs. Measure: partial switch to vouchers redeemable with local traders, with the donor's agreement, and revision of the target to date in the next report.

An indicator tracking table without variance analysis is only good for reporting. It is the "explanation and corrective action" column, discussed in a meeting and followed up at the next one, that turns it into a management tool.

Methodological note: for an indicator that is supposed to decrease (such as the rCSI), the achievement percentage is not calculated by simple division. Use the share of the expected change actually achieved: (baseline − achieved) / (baseline − target) = (21.4 − 13.1) / (21.4 − 14) = 112%. Specifying this rule in the MEAL plan avoids inconsistent tables from one report to the next.

8Writing an operational MEAL plan: a complete template

The MEAL plan is not a proposal document: it is the operating manual of the system for the team. It must be short (15 to 25 pages, excluding annexes), known by the team and revised at least once a year or whenever the project changes.

Standard outline

1. Introduction: project summary, objectives of the MEAL system, data users (team, management, partners, donor, communities, cluster).
2. Theory of change and logframe, with the priority learning questions (3 to 5 at most).
3. Indicator matrix: for each indicator, the definition, method, frequency, person responsible, baseline and target; the full reference sheets in an annex.
4. MEAL calendar: surveys, monitoring, reviews, reports, evaluation.
5. Methods and tools: description of each tool and its use, sampling for each survey.
6. Data flow and data management: from the point of collection to the decision, with deadlines.
7. Data quality assurance: checks, verifications, data quality assessments.
8. Accountability: information, participation, complaints and feedback mechanism, procedure for sensitive complaints.
9. Learning and data use: review meetings, after-action reviews, capitalisation of experience, feedback to communities.
10. Roles and responsibilities, including those of partners.
11. Data protection and ethics.
12. MEAL budget and resources.

Standard MEAL calendar for a 12-month project

The data flow: a written example

1. Day D: distribution; the programme team registers beneficiaries in the digital form (signed distribution list or fingerprint).
2. D + 2: the MEAL officer checks 100% of the lists (duplicates, unregistered households, discrepancies with the validated list) and reconciles them with the logistics delivery notes.
3. D + 5: discrepancies are reported to the project manager, who explains them in writing.
4. The 5th of the following month: the MEAL officer updates the consolidated database and the ITT.
5. The 10th: the MEAL coordinator validates the ITT and writes a one-page analysis note.
6. The 15th: monthly programme and MEAL review meeting; decisions are recorded in an action log with a person responsible and a deadline.
7. Every quarter: simplified feedback session with community committees.

Indicative breakdown of a MEAL budget

9Common mistakes, donor expectations and context variations

What donors, auditors and evaluators look at

• Traceability: for a figure in the report, can you trace it back to the source documents (signed lists, forms, database)? Donor verifications and financial audits often pick a few figures at random and ask for the evidence.
• Consistency between narrative reports, financial reports and tracking tables: the number of kits distributed must match stock releases and purchases.
• Stability of definitions: an indicator whose method changes between baseline and endline no longer measures change. Any change must be documented and, if possible, the baseline recalculated.
• Actual disaggregation, not just promised disaggregation.
• Explanation of variances: exceeding a target by 150% is as suspect as underachievement; it often signals a poorly calculated target or double counting.
• Data quality assessment: the commonly used data quality standards (validity, integrity, precision, reliability, timeliness) come in particular from USAID guidance, which provides for a data quality assessment at least every three years for reported indicators. Many NGOs apply it annually to their key indicators.

Common mistakes

Variations by context

• Sudden-onset emergency: start with 5 to 8 output and coverage indicators aligned with the cluster and a weekly tracking table, and add outcome indicators when the project is revised. The cluster situation report (often weekly at the start) is a priority data user.
• Protracted crisis: invest in resilience and sustainability indicators (functionality of water points 6 or 12 months after rehabilitation, income, savings) and in comparability between successive projects.
• Camp: use site management population data as the denominator, checking when it was last updated.
• Urban setting: denominators are uncertain; prefer indicators measured on the beneficiary sample and explain population estimates.
• Remote management: reduce the number of indicators, favour those that can be verified by independent means (verification calls, photos, payment service provider data).
• Local partners: harmonise reference sheets with the partner from the outset, provide tools in its working language and practical training on filling in the ITT; do not require more indicators than the donor asks for.

10Interview questions and key references

Interview questions

1. "Write an outcome indicator for a hygiene promotion project and explain how you would measure it."

A good answer proposes a practice indicator rather than a knowledge indicator (for example "% of households with a handwashing facility with water and soap, observed by the enumerator"), gives the numerator and the denominator, the method (observation during a sample household survey), the frequency (baseline and endline), the disaggregation, and flags the biases (self-reporting bias, observation effect).

2. "Your project reports 25,000 beneficiaries, the cluster has recorded 14,000. How do you explain this?"

It immediately raises double counting between activities, indirect beneficiaries, differences in definitions between the NGO and the cluster, and proposes recalculating unique people from the database. It mentions the need to align definitions with those of the cluster.

3. "How do you set a realistic target?"

It draws on the baseline, standards (Sphere, cluster), the results of comparable projects, the project's resources and duration, and plans for revision after the baseline. It can explain the difference between an annual target and an end-of-project target, and between cumulative and non-cumulative indicators.

4. "What does a MEAL plan contain, and how do you know it is being used?"

It lists the main components, then focuses on evidence of use: review minutes with decisions, an action log, programme changes, feedback to communities.

5. "How do you measure disability in your surveys?"

It cites the Washington Group Short Set, the six domains, the "a lot of difficulty" cut-off, the need to word questions in terms of functional difficulties and to train enumerators, and the IASC Guidelines on Inclusion of Persons with Disabilities.

Key references

• The Sphere Handbook (2018) and its key indicators by sector.
• Humanitarian Indicator Registry: OCHA and the global clusters.
• Technical guidance on the Food Consumption Score, the Reduced Coping Strategies Index and the CARI approach: World Food Programme.
• SMART methodology manual and Global Nutrition Cluster guidance.
• Washington Group Short Set and Child Functioning Module (Washington Group and UNICEF).
• IASC Guidelines on Inclusion of Persons with Disabilities in Humanitarian Action (2019).
• USAID guidance on data quality standards and performance indicator reference sheets (ADS 201).
• BHA indicator handbook; ECHO key results indicators.