How to write SMART indicators for humanitarian projects
Check each indicator against five SMART questions, define numerators and denominators, and fix weak wording with before-and-after examples.
Short answer
A SMART indicator is specific, measurable, achievable, relevant and time-bound. Define exactly what you count, the unit, any groups to disaggregate by, and for percentages the numerator and denominator. Check that you can collect the data with your team and budget, that the target is realistic, and that the indicator really shows the change in your result.
Key points
- Test every indicator with the five SMART questions.
- Two staff members should read an indicator and understand it the same way.
- For percentages, define the numerator and the denominator.
- Check that you can collect the data with your time, staff and budget.
- Fewer, clearer indicators beat long lists.
What is an indicator?
An indicator is a clear signal that shows whether a change has happened. Catholic Relief Services' practical indicator guidance describes SMART indicators as ones that gather reliable evidence about how much or how well objectives are being achieved. They define the quantity and quality of expected change.
An indicator is not the result itself. If the result is "Households have safe water," the indicator might be "Percentage of target households using water from a protected source." The indicator tells you how to recognise the result.
What do the five SMART questions ask?
The CRS guidance gives these checks for each letter.
| Letter | Ask yourself |
|---|---|
| Specific | Is it clear what is measured? Would two staff understand it the same way? Are vague words defined? Is the unit stated, with disaggregation if needed? For percentages, are the numerator and denominator defined? |
| Measurable | Can the data be collected and analysed in time? Can it be observed, counted or self-reported? Do the time, staff and skills needed fit the project? |
| Achievable | Is the change achievable in the project's time and resources? If the data changes by season, can you collect it at the same time each year? Are targets set and realistic? |
| Relevant | Is it the best measure of this result? Will it help managers decide? Will it help test your pathway of change? |
| Time-bound | Does it say when the change is expected? Are targets set for clear periods? |
The guidance notes that relevance, achievability and measurability are often not obvious from the wording, so review the indicator next to the result it measures.
How do you write one in five steps?
- Start from the result. Copy the exact wording from your logframe.
- Choose what to count. People, households, events or documents? Choose the unit.
- Define every vague word. What counts as "trained", "improved" or "access"? Write the definition.
- Add the numerator and denominator for any percentage. Numerator: households using a protected source. Denominator: all surveyed target households.
- Add groups and timing. Decide on disaggregation, such as sex, age or disability (see disaggregated data), and when you will measure.
What do weak and strong indicators look like?
| Weak | Why it fails | Stronger |
|---|---|---|
| "Improved hygiene practices" | Not specific or measurable | "Percentage of surveyed households with a handwashing station with soap and water within 10 metres of the latrine" |
| "Number of trainings" | Counts activity, not change | "Percentage of trained volunteers who score 70% or more on the post-training test" |
| "Community satisfaction" | No unit, no definition | "Percentage of respondents who say they are satisfied or very satisfied with the distribution, by sex" |
| "Reduce malnutrition in six months" | Not achievable in the period | Choose a shorter-term result such as feeding practices |
These examples are illustrations. Adapt thresholds to your project and, where possible, to recognised standards such as Sphere.
How do you set targets?
Use the baseline, your resources and experience from similar projects. If the baseline is 32% and you have one year, ask what is realistic with your budget. Ambitious targets that you miss can make a good project look like a failure. Record how you chose the target, so the next person understands it. See our guide on baseline surveys.
Who should help write the indicators?
Do not write indicators alone at a desk. CRS guidance says indicator development should involve the people who will use and collect the data. Ask programme staff what they can observe. Ask field staff what is practical to collect. If possible, ask community members what change matters to them. This catches indicators that look good on paper but cannot be collected in the field, and it makes staff more willing to collect the data carefully.
What mistakes should you avoid?
- Words that hide meaning. "Capacity", "awareness" and "effective" need definitions.
- Counting outputs as outcomes. The number of people trained is an output. What they do afterwards is an outcome.
- More than one thing in one indicator. Split "Percentage of women who attend and apply" into two.
- Indicators that nobody can collect. In areas with limited access, check that you can collect the data safely. See remote monitoring.
- No written definition. Put each definition on an indicator reference sheet.
How EMPO can help
EMPO is a reliable Myanmar-based training and consultancy provider and one of the country's leading names in practical NGO training, with 1,200+ people trained and 15 partner organisations. Our MEAL training covers indicators, data collection and reporting in English and Burmese, online or in person across Myanmar. Teams can also commission support to review their indicators. See our services or contact us.
Sources
- Practical Guidance on Developing Indicators — Catholic Relief Services, 2022
Questions people ask
Should an indicator include the target?
Follow your donor's template. Some donors want the target in the indicator wording, and others want the indicator written neutrally with baseline and target in separate columns.
Can an indicator measure something qualitative?
Yes. A qualitative change, such as satisfaction or practice, can be reported as a number, such as the percentage of respondents who say they are satisfied, if you define the terms and collect the data in a consistent way.
How many indicators should a project have?
Use as few as give a clear picture of each result. Every indicator costs time and money to collect, so remove those you will not use.