How to write SMART indicators for a donor logframe
An indicator tells the donor how you'll know a result has been achieved. Weak indicators are one of the most common reasons a proposal loses points, and one of the easiest to fix.
What SMART means in practice
- Specific: it names exactly what changes and for whom.
- Measurable: it has a number, a percentage or a clear yes/no.
- Achievable: the target is realistic for your budget and timeframe.
- Relevant: it measures the result it sits under, not a different one.
- Time-bound: it says by when.
In practice, a good indicator has five parts: who, what change, how much, by when, and measured how (the means of verification).
Before and after
| Weak indicator | SMART version |
|---|---|
| Improved awareness of hygiene | % of targeted households in 12 union councils who can name three critical handwashing times, from 20% at baseline to 70% by month 12 (household survey) |
| Women economically empowered | Number of women trained who report a monthly income of at least PKR 15,000 six months after training: target 300 of 500 (follow-up survey) |
| Better quality education | % of grade 3 students in 40 target schools reading a grade-level text with comprehension, from 30% to 55% by end of project (ASER-style reading assessment) |
| Capacity of partners built | Number of partner organisations scoring "adequate" or above on all five areas of the capacity assessment tool, from 1 to 5 of 6 by month 18 |
Match the indicator to the level
A logframe has levels, and each level needs a different kind of indicator.
- Impact (goal): long-term change the project contributes to. Usually drawn from national or donor data, such as district literacy rates. You aren't solely responsible for it.
- Outcome: the change in behaviour, practice or condition you are accountable for. This is where most scoring happens.
- Output: what the project directly delivers, such as people trained, schools rehabilitated or kits distributed.
- Activity: usually not given indicators; these are tracked in the workplan and budget.
A common mistake is placing output indicators ("500 teachers trained") at outcome level. Training is something you deliver. The outcome is what the teachers then do differently.
Always include a baseline and a target
An indicator without a baseline can't show change. If you don't have baseline data yet, say how and when you'll collect it ("baseline to be established in month 2 through a household survey") and give an indicative target. Donors accept this far more readily than a missing baseline.
Disaggregate
Most donors expect indicators broken down by sex, age and, where relevant, disability. Write it into the indicator: "% of trained youth (disaggregated by sex and disability) who…". For disability, the Washington Group Short Set of Questions is widely accepted and easy to add to surveys.
Means of verification: be concrete
"Project reports" is not a means of verification. Name the actual source: attendance registers, pre and post tests, household survey, school enrolment records, facility registers, photos with GPS, partner financial reports. If a donor can't picture how you'll collect it, they'll doubt it.
How many indicators?
Fewer, better indicators are easier to score and easier to report on. As a rough guide, aim for two or three indicators per outcome and two to four per output. Every indicator you add is data you'll have to collect, clean and report for the life of the grant.
Quick checklist
- Does each indicator say who, what change, how much, by when and how measured?
- Is every outcome indicator measuring change, not delivery?
- Does every indicator have a baseline (or a plan to get one) and a target?
- Are targets realistic for the budget and timeline?
- Is data disaggregated by sex, age and disability where it matters?
- Can your team actually collect this data with the budget in the proposal?
Next: turn these indicators into a MEAL plan.
