What a logic model is
A logic model is a simple diagram of the cause-and-effect logic behind your program. It answers one question for a funder or board member at a glance: if you put these resources in and do these activities, what changes, and why? Done well, it makes a program read as designed rather than improvised — and it keeps your own team honest, because any activity that doesn't lead to an outcome stands out immediately.
The five parts, defined
- Inputs — the resources you invest: staff, funding, volunteers, space, partners, curriculum.
- Activities — what the program does with those inputs: the workshops, coaching, meals, services.
- Outputs — the countable products of those activities: people served, sessions delivered, meals distributed.
- Outcomes — the changes that result, usually in knowledge, behavior, or condition: short-term, then medium-term.
- Impact — the long-term change you exist to create, often shared with other organizations.
The line that trips everyone up
Outputs are what you did (24 workshops). Outcomes are what changed (70% gained a job). Counting outputs and calling them results is the single most common logic-model mistake — and funders notice.
A full worked example
Here's a logic model for a fictional after-school literacy program serving struggling third-graders.
| Inputs | Activities | Outputs | Outcomes | Impact |
|---|---|---|---|---|
| 2 part-time tutors; reading curriculum; donated classroom; $40k program budget; school partnership | Twice-weekly small-group tutoring; weekly take-home reading; pre/post assessments | 60 students enrolled; 1,200 tutoring sessions delivered; 60 reading kits sent home | Short: students read 20+ min/day at home. Medium: 80% gain ≥1 reading level by June | More local children read at grade level, improving long-term school success in the district |
Notice how each column feeds the next: the inputs make the activities possible, the activities produce the outputs, the outputs drive the outcomes, and the outcomes accumulate into impact. Nothing appears that doesn't connect.
How to build one (step by step)
Read a logic model left to right, but build it right to left — starting from the change you want keeps every activity tied to a purpose.
- Start at the end with your impact. Name the long-term change you exist to create.
- Define the outcomes that lead to it. List the short- and medium-term changes in knowledge, behavior, or condition that drive that impact.
- List the outputs each activity produces. The countable products — people served, sessions delivered.
- Map the activities. Describe what your program actually does to produce those outputs.
- List the inputs you need. The staff, funding, partners, and materials the activities require.
- Check the chain reads cleanly. Follow it left to right — every link should connect with no orphaned activities or unsupported outcomes.
Why it's the backbone of grant proposals
Many grant applications ask for a logic model outright, and even when they don't, it's the skeleton of a strong methods section. It hands a reviewer the entire causal chain on one page and proves your outcomes aren't wishful — they're the visible end of a connected design. Carry your logic model straight into your grant proposal, and pair it with solid grant-ready program design and impact reporting.
A steady funding base is the input that makes the rest possible.
Every logic model starts with inputs — and funding is the one that quietly limits all the others. Good Circles gives your nonprofit recurring, unrestricted income with almost no labor: supporters pick your cause once, then a share of their everyday local spending funds you automatically (≈ $36,000/year from 500 supporters, free to join). It's the kind of dependable input that lets you run the program your logic model describes.
See how it works for nonprofits →Sources & tools
Free first
- W.K. Kellogg Foundation — Logic Model Development Guide (PDF) — The definitive free guide with checklists and templates for building inputs → activities → outputs → outcomes → impact.
- BetterEvaluation — Develop Programme Theory / Theory of Change (Rainbow Framework) — Methods and examples for constructing a logic model / program theory and avoiding common gaps.
- CDC — Logic Model resources (Program Evaluation) — Government-grade explanation and examples of logic models as the backbone of an evaluation plan.
- University of Wisconsin–Extension — Logic Model resources — Classic, widely cited free curriculum, templates, and worked examples for building program logic models.
- Candid Learning — Logic Models and Program Planning — Beginner-friendly explainer connecting the logic model to proposals and evaluation.
Paid — optional labor-savers
- Miro — Collaborative whiteboard for building and revising logic models with a team in real time. Worth it when Staff and board co-create the logic model and you want a live, shareable visual artifact.
- Lucidchart — Diagramming tool with logic-model and flowchart templates that export cleanly into proposals. Worth it when You need polished, presentation-ready logic-model diagrams for grant applications.
Last verified 2026-06-16. Figures and rules change — verify at the source before you act.
FAQ
What is the difference between outputs and outcomes?
Outputs are the countable products of your activities, such as 200 people served or 24 workshops delivered. Outcomes are the changes that result, such as 70% of participants gaining stable housing. Outputs prove you did the work; outcomes prove the work mattered.
Should I build a logic model left to right or right to left?
Read it left to right (inputs to impact), but build it right to left. Start with the impact you want, define the outcomes that lead there, then work back to the activities and inputs needed. Building backward keeps every activity tied to a purpose.
Do funders actually want a logic model?
Often yes — many applications ask for one directly, and even when they don't, it's the backbone of a strong methods section. It shows reviewers the causal chain from resources to results on a single page, which reads as designed rather than improvised.