Impact indicators translate changing the lives of others into clear metrics to decide. A good indicator not only describes, it guides actions: it shows direction, magnitude and pace of change. To be useful, they must arise from the theory of change of your project: first you define what change you want to see and then you decide what to measure. This means that not all “manual” or international indicators are suitable for every case. The challenge is to choose and adjust those that make sense for your intervention, and distinguish between outputs (what happens immediately), results (what takes time to happen) and impacts (what changed in the lives of others). Here you will learn to design, calculate and report them without confusion.
What are they for (and what are they not)?
An impact indicator serves to make decisions: adjust delivery, reallocate resources, scale or close. It is not a pretty figure for the final report, if it does not trigger a possible action, it is decorative.
To put ourselves in common language: outputs are what happens immediately (workshops given, attention), results are what takes time to happen (skills, behavioral changes, income), impacts are what changed in the lives of others (sustainable well-being, autonomy, stable employment). The impact indicators observe this last level, without losing sight of the previous ones.
Types and levels: how impact is connected to the chain
The logical chain helps you locate the impact indicator within the project and avoids asking “impossible things” from it. This comparison between levels serves to align expectations.
| Level | Answering question | Indicator example | Typical decision |
|---|---|---|---|
| Output | Was what was promised delivered? | % of appointments made over scheduled | Hedging operational adjustment |
| Result | Did anything change in the participants? | Soft skills score, monthly income | Refine content, intensity, channels |
| Impact | What changed in people's lives? | Employment within 6-12 months, reduction of avoidable hospitalizations | Scale, redesign, allocate budget |
Measure at all levels if you need to, but be clear that the protagonist here is the impact indicator: it is the one that tells you if people's lives are changing as you promised.
How to design them well, without turning them into a labyrinth
Start from the theory of change and land your indicators like this
- Measurable (viable): An impact indicator must be able to be calculated with the data, time and budget you have today, not with what you would like to have one day. “Measurable” also speaks of ease: that it is easy to build in the field, to register in the systems that already exist and to update it constantly.
- Reliable: A reliable indicator is one that can be believed because it is calculated with clear and stable rules, using clear and consistent sources. Your numbers are consistent with what we know about the program and the context, and anyone on the team, with the same rules and data, can recalculate it and get the same result.
- Communicatable: a useful indicator can be explained in a simple sentence to management, teams and allies, without having to give a complete class every time it appears in a report. If no one remembers what it means or how to read it, it will be difficult to use it to decide.
- Focused: A good indicator points to the key change you want to see, it doesn't try to measure everything at once. Avoid “Christmas tree” indicators with too many things hanging (three dimensions, four conditions, several times in one). Better a clear indicator per important idea than a confusing one that mixes many.
- Relevant/coherent: the indicator must make sense within your theory of change and the actual project decisions. It can rely on literature or international standards, but not be there just “because it is what everyone uses.” It has to talk with the context, the population and the type of decision: if you never look at it to change anything, it is probably not relevant.
In practice, there is no perfect indicator that wins on all five criteria at once. There is always a balance between how accurate it is, how much it costs to measure it, how often you can update it and how easy it is to use and explain. That is why it is worth starting with a phase of ideation and search for references (literature, standards, other organizations) and then prioritize and ground those indicators to your capabilities, your context and your theory of change.
Calculation and reading: units, percentages and percentage points
Impact indicators are usually summarized as rates or averages. Two rules avoid confusion:
- Unit change = difference in real units (pesos, days, points).
- Percentage change = the difference over the initial value × 100. For rates, clarify if you are talking about percentage points (direct difference between percentages).
Simple example. Youth employment: insertion rate goes from 25% to 35%.
- Change in percentage points: +10 p.p.
- Percentage change: +40% (10/25).
- Decision: deepen the mechanism that worked in the segment where it rose the most.
Example in health. Rehospitalizations within 90 days drop from 14% to 10%.
- Change: −4 p.p. and −28.6%.
- Decision: consolidate the home monitoring component.
Common mistakes (and how to avoid them)
- Confusing reach with impact: reaching 10,000 people is not changing 10,000 lives, it is always accompanied by result/impact indicators.
- Ambiguous formulations: “improve well-being” is diffuse, uses concrete scales anchored in time (“last 30 days”).
- Change the rule of the game: do not modify the formula, scales or period in the middle of the year, if you must do so, document the reason and break series clearly.
- Set goals without basis: estimate goals with past data or pilots, avoid impossible or trivial promises.
- Averaging percentages without context: Instead of taking a single average of many percentages, it is better to calculate the percentage using the sum of all the cases together or show the percentages by subgroup saying how many people are in each one.
- Forget seasonal bias: compare equivalent periods (same quarter of the year), take care of holidays, harvests, school cycles.
Examples by sector (to land ideas)
Employment and training.
- Impact: formal employment rate at 6/12 months.
- Result: completion of practical training, intensity of active search.
- Output: tutoring sessions carried out.
Community health.
- Impact: reduction of avoidable rehospitalizations, healthy days.
- Result: adherence to treatment, control of warning signs.
- Output: home visits completed.
Basic education.
- Impact: annual school permanence, transition to secondary education.
- Result: progress in reading/mathematics (standardized tests).
- Departure: effective class hours.
Atmosphere.
- Impact: reduction of critical discharge points, water quality.
- Result: adoption of recycling practices, recovered volume.
- Departure: workshops and collection operations.
How to connect them with real decisions?
The indicators must “speak” with the operation. Practical rules:
- Weekly/biweekly: watch outputs to ensure coverage and quality.
- Monthly/bimonthly: reviews results and corrects design (content, schedules, channels).
- Quarterly/semi-annually: evaluates impact progress and decides to scale, adjust or pause.
In addition, it defines action thresholds: if adherence falls <70%, a visit is activated, if female employment < goal, reinforces support for care.
Measuring without deciding is useless.
Frequently asked questions
/ 10
Conclusion
Impact indicators are the compass to fulfill your promise to social impact -change the lives of others- with evidence and focus. Designing them well avoids eternal discussions and speeds up decisions: what to keep, what to adjust, what to scale. Do you want to debug your set of indicators and connect it with your dashboard and evaluation? Let's talk.