A theory of change template with a labeled box for every part - plus an AI wizard that drafts the whole framework from a description of your program.
A theory of change template is a set of empty boxes — problem, activities, outcomes, impact — and the boxes were never the hard part. The hard part is knowing what belongs in each one and whether the logic holds. The Sopact theory of change wizard starts from a plain-language description of your program, drafts every box, names the assumptions between them, and connects each outcome to the data that would test it. You still get the template — you just start from a draft, not a blank canvas.
By Unmesh Sheth · Founder & CEO, Sopact · Updated May 2026
A theory of change template is a pre-structured framework with a labeled space for each part of a theory of change — the problem, the activities, the outcomes, the long-term impact, and the assumptions that link them. A template gives a team the structure to fill in. Whether the theory holds depends on what goes in each box — and whether each outcome can be tested with real data.
Theory of change template, theory of change canvas, theory of change builder — the same need: a fast way to a first draft. This page gives you the template and the AI wizard that fills it; for the framework behind it, see the theory of change guide.
A blank template asks you to know the answer before you start. The wizard works the other way — edit the description below, press Generate, and a drafted theory of change appears. Then click any box to make it yours.
Edit any box above, then copy your draft. To build the full version — every outcome wired to live data — book a working session.
A paragraph: who you serve, what you do, the change you expect. No framework, no jargon — the wizard handles the structure.
Problem to impact, with a named assumption on every link — a complete first draft, not an empty canvas.
Click any box or assumption and rewrite it. The structure stays; the words become yours.
Copy the draft for a proposal — or build the full version in Sopact, where every outcome is connected to the data that would test it.
The wizard does not decide what your program is for — you do. It removes the blank-canvas problem: you start from a draft specific enough to argue with, edit it in place, and leave with a theory of change instead of an empty page.
Filling it in by hand? Here is the whole template — every box with the question it answers and a one-line example. The wizard answers these from your program description; either way, the structure is the same.
| Box | The question it answers | A one-line example |
|---|---|---|
| Problem | Who is affected, and what causes it? | Young people face 35% unemployment with few routes into tech. |
| Inputs | What do you commit before activities start? | Instructors, a curriculum, employer partners, equipment. |
| Activities | What does the program actually deliver? | A 12-week cohort bootcamp with one-to-one mentoring. |
| Outputs | What countable products come out? | Completed portfolios and earned credentials. |
| Outcomes | What changes for the people you serve? | Graduates placed in tech roles within six months. |
| Impact | What long-term change do you contribute to? | Durable income gains across the cohort. |
| Assumptions | What must be true for one box to lead to the next? | Employers value the credential; hiring volume holds. |
Read it backward — from impact to problem — and check the logic holds at every link. And give every outcome a data instrument; an outcome you cannot measure is decoration on the canvas. The wizard checks both on every box.
Filling it in by hand? Here is the whole template — every box with the question it answers and a one-line example. The wizard answers these from your program description; either way, the structure is the same.
| Box | The question it answers | A one-line example |
|---|---|---|
| Problem | Who is affected, and what causes it? | Young people face 35% unemployment with few routes into tech. |
| Inputs | What do you commit before activities start? | Instructors, a curriculum, employer partners, equipment. |
| Activities | What does the program actually deliver? | A 12-week cohort bootcamp with one-to-one mentoring. |
| Outputs | What countable products come out? | Completed portfolios and earned credentials. |
| Outcomes | What changes for the people you serve? | Graduates placed in tech roles within six months. |
| Impact | What long-term change do you contribute to? | Durable income gains across the cohort. |
| Assumptions | What must be true for one box to lead to the next? | Employers value the credential; hiring volume holds. |
Read it backward — from impact to problem — and check the logic holds at every link. And give every outcome a data instrument; an outcome you cannot measure is decoration on the canvas. The wizard checks both on every box.
A downloaded canvas and an AI wizard both end in a theory of change. They differ in where the effort goes — and in whether the theory can be tested.
| The decision | A blank template | The theory of change wizard |
|---|---|---|
| Starting point | Eight empty boxes you fill from memory | A drafted theory of change from your program description |
| Assumptions | Easy to skip — the boxes do not ask for them | Named on every link, where the logic is riskiest |
| Outcomes and data | Listed as text, with no instrument attached | Each outcome wired to the instrument that tests it |
| Time to a first draft | A blank-page afternoon, or a workshop | Minutes — then you edit |
| After the draft | A file that goes in a drawer | A living canvas you revise as evidence arrives |
A blank template is fine if you only need the shape on a page. The wizard is for the part the shape cannot do — getting the content right, and keeping the theory testable once the program is running.
A short walk-through of the sequence the wizard follows: draft the framework fast, connect every outcome to a collection instrument, and revise the theory as evidence arrives — instead of finishing the template and filing it.
Designing a theory of change against data, not before it. The full framework is in the theory of change guide.
A theory of change template is a pre-structured framework with a labeled space for each part of a theory of change: the problem, inputs, activities, outputs, outcomes, long-term impact, and the assumptions that link each step. The template gives a team the structure to fill in. It does not, on its own, tell you whether the content in each box is right or whether each outcome can be measured — which is the part that decides whether the theory holds.
Yes. ActKnowledge, the Annie E. Casey Foundation, and the W.K. Kellogg Foundation all publish free theory of change templates, and the structure is broadly the same. This page reproduces the full template with the question each box answers and a one-line example, so you can fill it in directly. The harder question is not where to find a blank canvas; it is what belongs in each box — which is what the AI wizard is for.
A theory of change builder, or generator, is a tool that produces a theory of change rather than handing you a blank template. The Sopact wizard is one: you describe your program in plain language, and it drafts every box, names the assumptions between them, and connects each outcome to a data instrument. A builder turns the blank-canvas problem into an editing problem — you start from a draft specific enough to correct, rather than an empty grid.
Work backward, not forward. Start with the long-term impact you contribute to, then the outcomes that lead to it, then the outputs, activities, inputs, and the problem at the base. At each link, write the assumption that has to be true for one box to lead to the next. Then give every outcome a data instrument — the survey, assessment, or record that will measure it. Read the canvas backward once more to confirm the logic still connects.
A standard theory of change template has six content boxes plus an assumption layer: the problem (who is affected and why), inputs (what you commit), activities (what you deliver), outputs (the countable products), outcomes (the change in the people you serve), and impact (the long-term change you contribute to). The assumptions run underneath — one for each link, naming the external condition that must hold for the logic to work.
Yes. Given a plain-language description of a program — who it serves, what it does, the change it expects — an AI wizard can draft a complete theory of change in minutes: every box filled, the assumptions named, and outcomes mapped to data. What AI should not do is finalize the theory unedited. The draft is the starting point; the judgment about what your program is for, and which assumptions are the risky ones, stays with the team.
A template is the empty structure — the boxes and the labels. A finished theory of change is that structure filled with a specific program's logic, with assumptions named and outcomes connected to data that can test them. The template is the easy part; many teams produce a tidy one and still cannot answer a funder's question about evidence. The full framework, and what makes a theory of change testable, is covered in the theory of change guide.
This guide is educational and based on publicly available theory of change methodology. Information current as of May 2026. To suggest a correction, email unmesh@sopact.com.
The full guide — the six-component framework, the assumptions layer, and the AI-age way to build one without losing six months.
Worked theories of change across sectors — each one filled in, with the assumptions named, so you can see the template in use.
How a theory of change and a logic model differ — what each one controls, and when a funder wants which.
A working session, not a demo. Tell us what your program does in plain language and we will run the wizard live — every box drafted, the assumptions named, and each outcome wired to the data that would test it. You leave with a theory of change you can put in a proposal and still revise against evidence.
Live walkthrough · with Unmesh Sheth, Founder & CEO · bring a program to describe
