Sopact is a technology based social enterprise committed to helping organizations measure impact by directly involving their stakeholders.
Copyright 2015-2026 © sopact. All rights reserved.
Four complete theory of change examples - education, workforce, public health, and environment - each a filled six-box canvas from problem to impact.
A theory of change example is a worked version of the six-component pathway — problem, inputs, activities, outputs, outcomes, impact — filled in for a real program, with the assumption named under every link, so you can see what a testable theory of change looks like. It turns an abstract framework into a concrete model you can adapt to your own program.
Most examples online show the boxes but skip the assumptions, and the assumption under each arrow is the part data tests. A strong example names the riskiest assumption at every link, not just the chain of results. That is the difference between a diagram you can file and a model you can measure against. See the underlying framework in the theory of change guide.
Used by: nonprofit program designers, grant writers, foundation officers, and M&E leads who need a worked model to adapt rather than a blank template to stare at.
Copy a typical online example and you get five boxes in a row: training leads to skills leads to jobs. It reads clean and it teaches nothing, because the moment your program differs — a different population, a weaker labor market, a shorter dosage — the borrowed chain no longer holds and the example gives you no way to know. The boxes travel; the reasoning does not.
The reasoning lives in the assumptions. Under every arrow sits a claim that must be true for the link to hold: that participants who gain skills will actually apply for the jobs, that employers value the credential, that the follow-up support reaches the people who need it. A worked example is only useful if it names those claims, because the assumption is the part your data can test and revise. This is the shift Sopact is built around: an example is not a picture to admire but a set of testable claims wired to indicators on a persistent stakeholder ID, so each outcome is measured against a real assumption rather than asserted. Sopact turns a static example into a tested theory of change on your own data.
Each example below runs the same spine — the problem, the key activity, the main outcome with its indicator, and one named assumption — so you can see the pattern hold across very different programs. Read them as scaffolds to adapt, not templates to copy.
Workforce training (a nonprofit coding program). The problem: unemployed adults in a mid-size city lack the technical skills local employers hire for. The key activity is a 12-week full-stack coding bootcamp with portfolio projects and interview coaching. The main outcome is stable employment in a tech role, and its indicator is the share of graduates placed in a job paying a living wage within 90 days, measured on a pre-post record. The riskiest assumption under the outcome arrow: that local employers actually value a bootcamp portfolio over a four-year degree. If hiring managers screen on credentials, skills gained never convert to jobs, so employer-hiring behavior is the first thing this theory monitors.
Education (a K-12 tutoring initiative). The problem: third-graders reading below grade level are on track to fall further behind each year. The key activity is one-on-one structured-literacy tutoring three times a week for a school year. The main outcome is reading at or above grade level by year end, with the indicator a pre-post gain on a standardized reading assessment scored on the same instrument at baseline and follow-up. The named assumption under the activity-to-outcome link: that tutoring dosage is high enough and consistent enough to matter. If attendance drops below roughly two sessions a week, the mechanism weakens, so session consistency is monitored as the leading signal, not just end-of-year scores.
Healthcare (a community diabetes-management program). The problem: patients with type-2 diabetes in an underserved area struggle to control blood sugar and cycle through avoidable ER visits. The key activity is a community-health-worker program pairing each patient with a coach for home visits and medication support. The main outcome is improved glycemic control, indicated by the change in HbA1c from enrollment to six months on one patient record. The riskiest assumption: that patients can access and afford the medication and diet the coaching recommends. If cost or supply blocks adherence, coaching alone will not move HbA1c, so a barrier check on affordability is embedded mid-program rather than discovered at the end.
Agriculture (a smallholder farmer training program). The problem: smallholder farmers use practices that leave yields and incomes low and vulnerable to drought. The key activity is season-long training in climate-resilient techniques plus access to improved seed. The main outcome is higher household income from farming, with the indicator the change in net crop income across one full season, tracked per household. The named assumption: that farmers can sell the additional harvest at a fair price. If market access or price collapses at harvest, higher yield does not become higher income, so the theory watches the market-access link as closely as the yield link. Reference layouts for any of these sit in the theory of change diagram guide.
Watch — worked theory of change examples. How a single example is built out component by component, with the assumption named under each link so the model can be tested rather than filed. Presented by Unmesh Sheth.
A worked example is the bridge between two other formats teams reach for. Strip the assumptions and mechanisms out of any example above and what remains — inputs, activities, outputs, outcomes, impact in a row — is a logic model: the operational summary a funder reads in under a minute. Add the mechanism on each arrow and the assumption behind each outcome and you are back to the full theory of change. The full contrast between the two is in the theory of change vs logic model guide.
To adapt an example to your own program, start from a blank canvas and fill the same six components with your problem, your activity, and your outcome. That structured canvas is the theory of change template. The example shows you what a filled-in pathway looks like; the template gives you the empty one to complete; and connecting each outcome to an indicator on a persistent ID is what turns the finished example into a live M&E system, covered in the theory of change in monitoring and evaluation guide.
An example earns its keep at four moments — picking a worked model close to your program, adapting it and attaching an indicator to each outcome, naming the assumption under each link and collecting on one ID, then testing and revising each cycle. The animation below runs the loop; the four prompts under it are the ones behind each job.
1 · Build the worked example. Draft the six components for your program, with the assumption named under every arrow rather than left implicit. The walkthrough is in how to build a theory of change.
Academy walkthrough → How to build a theory of change
Build a worked theory of change example for this program: [PROGRAM URL OR DOC]. Fill all six components - problem, inputs, activities, outputs, outcomes, impact - and for every outcome name the indicator that measures it and the single riskiest assumption under that link. Return it as a model I can test, not just a diagram.
2 · Audit the example for hidden assumptions. Find the arrows with no named assumption and the outcomes with no way to measure them, before a funder does.
Academy walkthrough → How to audit a theory of change
Audit this theory of change example: [PASTE OR LINK]. Flag every arrow with no stated assumption, every outcome that is really an output, and every outcome with no indicator. Rank the assumptions by how much the whole theory depends on them, and mark which one is riskiest.
3 · Derive the logic model. Strip the example down to the five-component operational summary a funder reads first.
Academy walkthrough → How to build a logic model
From this theory of change example: [PASTE OR LINK], derive the logic model beneath it: inputs, activities, outputs, outcomes, and impact, with an indicator on every output and outcome. Keep it consistent with the example rather than rebuilding from scratch.
4 · Convert to the funder's matrix. Turn the same example into the logframe a funder asks for, assumptions column included.
Academy walkthrough → How to build a logframe
From this theory of change example: [PASTE OR LINK], produce a logframe: goal, purpose, outputs, and activities, each row with its indicator, means of verification, and the assumption from the example. Keep the assumptions consistent with the ones already named.
The sections above are the worked models; the Academy articles are the practice - each a hands-on companion written to run on your own data.
Yes. A workforce example: the problem is that unemployed adults lack the skills local employers hire for; the key activity is a 12-week coding bootcamp; the main outcome is stable employment, indicated by the share of graduates placed in a living-wage job within 90 days on a pre-post record; and the riskiest assumption is that employers value a bootcamp portfolio over a degree. In Sopact each outcome in an example is wired to an indicator on a persistent stakeholder ID, so the assumption is tested rather than asserted.
A good theory of change example includes all six components - problem, inputs, activities, outputs, outcomes, impact - but its real value is the assumption named under every link, because the assumption is the part your data tests. Most online examples show the boxes and skip the assumptions. In Sopact a worked example attaches an indicator to each outcome and names the riskiest assumption at each arrow, so the example becomes a model you can measure against instead of a diagram you file.
A common nonprofit example is a coding program: the problem is that unemployed adults lack in-demand skills, the activity is a bootcamp with portfolio projects, the outcome is a living-wage job within 90 days measured on a pre-post record, and the riskiest assumption is that employers value the portfolio. In Sopact the nonprofit builds this example once and each outcome is collected on one participant ID from intake to follow-up, so the theory is tested on real data rather than described in a grant application.
An education example: the problem is third-graders reading below grade level; the activity is one-on-one structured-literacy tutoring three times a week; the outcome is reading at grade level by year end, indicated by a pre-post gain on the same standardized assessment; and the named assumption is that tutoring dosage stays high enough to matter. Sopact monitors session consistency as the leading signal, so a dosage problem surfaces mid-year rather than in the end-of-year report.
A healthcare example: the problem is patients with type-2 diabetes struggling with glycemic control; the activity is a community-health-worker coaching program; the outcome is improved control, indicated by the change in HbA1c from enrollment to six months on one patient record; and the riskiest assumption is that patients can access and afford the recommended medication and diet. In Sopact an affordability barrier check is embedded mid-program, so an adherence problem is caught while it can still be addressed.
An example is a filled-in pathway - a real program's problem, activity, outcome, and assumptions worked out - so you can see what a finished, testable theory looks like. A template is the empty canvas with a labeled slot for each component, waiting for you to fill it. Use the example to understand the pattern and the template to build your own. Sopact then wires each outcome to an indicator on a persistent ID, turning your filled-in template into a tested theory of change rather than a static diagram.
A theory of change example has six components: problem, inputs, activities, outputs, outcomes, and impact, read as a causal pathway from the need a program addresses to the long-term change it contributes to. What separates a strong example from a weak one is not the six boxes but the assumption named under each link between them. Sopact treats each of those assumptions as a testable claim tied to an indicator, so the example can be revised on evidence.
Start from an example close to your sector, then replace its problem, activity, and outcome with yours, keeping the same spine - one main outcome, its indicator, and the riskiest assumption under each link. Fill the six components on a blank canvas, which is what a theory of change template gives you. In Sopact you then attach each outcome to an indicator collected on one persistent stakeholder ID, so your adapted example is tested on your own data from the first cohort.