Practical evaluation guide

What is an evaluation indicator? Types, examples, development and differences from KPIs

Organize program and nonprofit evaluation indicators through types, examples and differences from KPIs. Includes a definition table for numerators, denominators, missing data and collection responsibilities, and fictional consultation-support entries. Use the free worksheet to specify your own indicators.

Turn questions into observable indicators

Main takeaway

An evaluation indicator expresses a change or implementation condition of interest in observable terms. Indicators can be qualitative descriptions as well as numbers. A good indicator corresponds to an evaluation question, defines the period, what is being judged and how data will be obtained, and specifies the numerator and denominator for proportions, so that it can inform decisions after collection. Keep targets describing the desired level separate. [1]

An evaluation indicator expresses a change or implementation condition of interest in observable terms. It concerns what to measure and is treated separately from a target defining the desired level.

Start by deciding the questions the evaluation should answer, rather than listing indicators. Adding indicators while questions remain vague can produce data without enabling judgment. This guide covers indicator types, development steps, characteristics of good indicators and practical selection checks. It explains evaluation of programs run by municipalities, nonprofits and similar organizations.

Create indicators that fit the question

Distinguish types of indicators and specify what to observe and how to measure it.

Indicator types and development from questions

Indicator types differ according to the program stage being examined. The table below is this site's practical classification. Names and categories vary by scheme and grant form, so prioritize the recipient's terminology.

After understanding the types, develop indicators from evaluation questions. If the question is whether training changed participants' practice, satisfaction is insufficient. Candidates include how often participants practiced afterward, products they created and sharing with supervisors or colleagues.

Conversely, if the question is whether the program is reaching the intended group as planned, process indicators such as participation rates, target characteristics, attrition and delivery counts become important.

Confirm the evaluation purpose
Write evaluation questions
Describe the changes or conditions to measure
Turn them into observable candidate indicators
Decide data sources and collection frequency
Set judgment criteria or reference values

Indicator types and development from questions
Type	What it examines	Example	Main data sources
Input indicators	Resources used	Budget expenditure, person-days and number of teaching materials	Accounting and operational records
Output indicators	What was done or provided	Training sessions, participants and consultations	Implementation records and registration logs
Process indicators	Implementation processes and reach	Participation rates, attrition and target-characteristic breakdowns	Implementation records and participant lists
Outcome indicators	Changes for the target group	Understanding, proportion putting learning into practice and retention	Surveys, follow-up studies and administrative data
KPIs and performance indicators	Indicators particularly important for management	Monthly trends in key indicators	Select and define from the above

[1]

Differences between output and outcome indicators

This guide distinguishes indicators of delivery volume from indicators of changes for the target group. Consultation counts and material distribution counts establish how much was delivered; behavior after consultation and similar measures examine subsequent change.

The table below is this site's classification example. The OECD's 2023 glossary includes changes in knowledge and skills directly produced by activities within Outputs, so classification may differ from this table, which treats learning as a short-term outcome. Check the recipient's definitions.

Differences between output and outcome indicators
Type	What is examined	Example
Output indicators	What was done or provided	Training sessions, participants and consultations
Initial outcome indicators	Changes in understanding, awareness and confidence	Understanding tests, self-efficacy and knowledge scores
Medium-term outcome indicators	Changes in behavior and practice	Frequency of practice, products created and actions after consultation
Long-term outcome indicators	Changes in conditions and results	Employment retention, learning attainment and organizational improvement

[2]

Characteristics of good indicators

Good indicators can inform decisions as well as be measured. Selecting them solely because numbers are easy to obtain can move the evaluation away from the changes of real interest.

Limit indicators to those needed to answer the questions. There is no fixed appropriate count. Test who records information and how many minutes it takes, and review indicators unused in decisions or excessively burdensome.

Directly corresponds to an evaluation question
Clearly identifies what changes and for whom
Data sources and collection methods are feasible
Can inform judgment or improvement after collection
Distinguishes short-, medium- and long-term horizons

[1]

Indicator definition table: align measurement conditions

Where there are multiple candidates, imagine how evaluation results will be used rather than considering ease of measurement alone. It may be necessary to exclude from the main indicators numbers that cannot inform meeting decisions, do not lead to improvement, or are too costly to collect.

Quantitative indicators alone may not reveal why change occurred. Combining main outcome indicators with qualitative information such as open responses, brief interviews and case records makes interpretation and consideration of improvements easier.

Indicator names alone do not ensure comparable aggregation. The table below contains entries proposed by this site. Specify calculation conditions for quantitative indicators and what to record and how to judge it for qualitative indicators. Do not fill gaps by guessing; assign responsibility and timing for checking them.

The indicator can answer the evaluation question
Staff use consistent data definitions
Collection frequency and responsibility assignments are feasible
Improvement actions can be anticipated after reviewing increases or decreases

Indicator definition table: align measurement conditions
Item to define	Fictional consultation-support entry	What to check
Evaluation question	Did people proceed with necessary applications after consultation?	Which decision will the results inform?
Target group and period	Check on the 20 people consulted during the target period 1 month after consultation	Fix inclusion rules and the checking timepoint
Indicator and counting unit	Number starting at least 1 necessary application	Which record confirms that an application started?
Numerator and denominator	9 confirmed starters / 15 people with known status	Distinguish all 20 targets from the 15 with known status
Missing data and exclusions	Report the 5 unconfirmed people separately rather than classifying them as starters or non-starters	Do not silently exclude unconfirmed cases
Data sources and collection responsibility	Consultation staff check follow-up records obtained with consent	Check participant burden and information handling
Judgment criteria and targets	Agree these with stakeholders beforehand and keep them separate from observed results	Do not turn the observed result of 9 people into a target
Use and limitations of results	Check barriers to starting applications and review support	Do not infer a causal effect of consultation from the proportion alone

Practical framework or suggestion

Fictional example; not actual results

Make progress toward applications countable

A fictional example for illustration: define delivery volume and behavioral change in follow-up of consultation support.

Item	Example entry	Reason
Indicator	Number starting an application within 1 month of consultation	Fix the period between consultation and checking
Denominator	20 follow-up targets, 15 with known status	Keep the 5 unconfirmed cases visible
Illustrative result	9 of 15 started	60% of those with known status
Supplementary information	Classify reasons people could not start	Consider improvements in light of changes in counts
Proportion with known status	15 of 20, or 75%	Also state the scope for which status is known

What this example supports

Do not treat people whose status is unconfirmed as zero or remove them from the denominator without explanation. Keep targets and observations in separate fields. The 60% from 9÷15 is the proportion among people with known status. The proportion of all 20 targets confirmed as starting is 9÷20, or 45%, but with 5 unconfirmed people it is not the actual starting proportion for the entire target group. These numbers alone also cannot establish the effect of consultation.

Common errors in judgment

Use easily measured satisfaction as the only main indicator

How to revise: If application behavior is the concern, check starting status and barriers

Reason: This prevents mismatch between questions and indicators.

Write only utilization rate and let staff use different denominators

How to revise: Define the period, numerator, denominator, exclusions and missing-data treatment

Reason: The same name can otherwise represent different numbers.

Put it into practice next

Record measurement methods, judgment criteria, missing-data treatment and collection responsibility for one indicator

View an example and open the worksheet Read the method explanation

Frequently asked questions

What is an evaluation indicator?

An item expressing changes or implementation conditions of interest in observable terms. It can be a qualitative description as well as a number. An indicator specifies what to measure and is kept separate from a target specifying the desired level.

Are KPIs and evaluation indicators the same?

They overlap but are not identical. KPIs are key management indicators, while evaluation indicators provide evidence for answering evaluation questions.

Is satisfaction an outcome indicator?

Satisfaction is useful supplementary information but weak evidence of outcomes on its own. Combine it with changes in understanding, behavior and conditions.

How many indicators are needed?

There is no universally appropriate count. Select indicators needed for the main questions and adjust the volume so that those responsible can sustain collection and analysis.

References and checked sections

CDC Program Evaluation Framework, 2024
Centers for Disease Control and Prevention · 2024
Section consulted
Steps 3–4: Focus the Evaluation Questions and Design / Gather Credible Evidence
How it relates to the text
Alignment and feasibility of purposes, questions, indicators and collection methods
Source checked: 2026-09-13 (includes AI cross-checking)
Glossary of Key Terms in Evaluation and Results-Based Management for Sustainable Development (Second Edition)
OECD · 2023
Section consulted
Outcomes / Outputs (main text, p.45)
How it relates to the text
Definitions of outputs and outcomes, and differences between frameworks in classifying changes in knowledge and skills
Source checked: 2026-09-13 (includes AI cross-checking)

Content updated: 2026-10-12Source checked: 2026-09-13 (includes AI cross-checking)No documented human reviewIncludes fictional examplesEditorial policy and scope of checking Cite this page Report a correction

How to cite this page

EvalCompass Editorial Team. “What is an evaluation indicator? Types, examples, development and differences from KPIs”. EvalCompass, updated 2026-10-12. https://evalcompass.jun-nakatani.com/en/guides/evaluation-indicators/

Add the date you accessed the page. Text created by this site is available under CC BY 4.0. Quotations, external materials, and bundled fonts have their own terms of use. When reusing content, credit the source, link to the license, and indicate any changes.

Suggest a correction or improvement to this page

Change history

2026-10-12 — Clarified the scope as program evaluation and added a measurement-condition definition table and examples distinguishing response rates from missing data.
2026-10-07 — A cross-guide review clarified indicator formats, outcome classifications and causal judgments, interpretation of tables, and the scope of SVI2026 where relevant.
2026-09-22 — Changed the introduction to begin with a definition of evaluation indicators. Added a table of indicator types and an FAQ on their definition.
2026-09-13 — Added links to original sources, scope of definitions, and fictional completed and judgment examples. Revised unsupported advice prescribing fixed counts or timing.

Turn questions into observable indicators