Skip to main content
[ Template library ]

Theory of Change template with a KPI set

From the problem you address to the change you claim, with an indicator attached to every step — so the framework is something you can measure against, not a diagram that lives in a grant proposal.

Use it when designing a program, answering a funder’s “how will you know” question, or deciding which twelve numbers your team should actually track.

[ The structure ]

The chain, then the numbers that prove each link

Illustrative examples · Synthetic sample data
01

The problem and the population

Who, where, and what is true for them now. Be specific enough that a stranger could recognize a participant.

Worked example · Youth Futures Academy

Students in grades 2–5 at five Phoenix sites reading below grade level at intake, with limited access to one-to-one reading support at home.

02

Inputs

What goes in: people, hours, money, instruments, partners. Quantify each — inputs are the denominator of every efficiency claim later.

Worked example · Youth Futures Academy

146 trained volunteer tutors · 5 host sites · Fountas & Pinnell assessment kits · $500K total program investment for the period.

03

Activities

What you actually do, with the dose: how often, how long, in what format. Dose is what your attendance data will later be compared against.

Worked example · Youth Futures Academy

One-to-one tutoring, twice weekly, fourteen weeks, with a consistent tutor per student.

04

Outputs

The countable products of activity — sessions, hours, people reached. Necessary, never sufficient: outputs prove you did the work, not that it worked.

Worked example · Youth Futures Academy

2,890 tutoring hours delivered · 1,214 students enrolled · 87% retained to week fourteen.

05

Outcomes

The changes in participants you can observe within the program’s horizon. Short-term and medium-term, stated as a direction plus a measure.

Worked example · Youth Futures Academy

Reading level rises (grade-level gain at exit vs intake) · reading happens outside the program (caregiver-reported, unprompted).

06

Impact

The longer-term change the outcomes are meant to add up to — named so the chain is visible, and clearly marked as beyond this period’s evidence.

Worked example · Youth Futures Academy

Grade-level literacy by the end of primary school. Not evidenced in a five-month period; stated as the destination, not a result.

07

Assumptions and risks

What has to be true for each arrow to hold, and what could break it. This is the section funders respect most and most organizations skip.

Worked example · Youth Futures Academy

Assumes tutor consistency drives attendance; risk that transport gaps at peripheral sites suppress dose below the threshold where gains appear.

08

The indicator set

For each outcome, two to four KPIs with a definition, a data source, a collection frequency, and a target. Twelve is plenty; forty is a spreadsheet nobody maintains.

Worked example · Youth Futures Academy

Outcome 01 · Reach and retention: students enrolled (intake form, monthly), retention to week 14 (attendance log, weekly), attendance by site. Outcome 02 · Reading level: F&P level at intake and exit, grade levels gained, gain by attendance band. Outcome 03 · Reading at home: caregiver-reported frequency (interview, end of cycle), books borrowed per student.

[ With Instrumento ]

The framework your whole workspace hangs off.

In Instrumento the Theory of Change is not a document — it is the structure every KPI, survey, dashboard, and report section is attached to, so nothing gets measured without a reason and nothing reported without a source.

Drafts are starting points. Your team reviews every figure and every sentence before anything reaches a funder.

See it filled in

Page three of the sample report shows this framework completed for Youth Futures Academy, with the indicator set tracked against each outcome.

View the sample report →

Start from the structure. Fill it with your evidence.

Bring one reporting challenge. We’ll show you how Instrumento drafts this template from your own data, with figures linked to their source rows.