Logic Model: How to Build One Step by Step

A logic model is a one-page visual that links five elements in sequence: inputs (the resources you invest), activities (what you do with them), outputs (the countable products of those activities), outcomes (the changes they produce in the people you serve), and impact (the long-term change you contribute to). In a grant proposal it shows a funder, at a glance, how money turns into results, and it ties your methods section directly to your evaluation plan. Every activity in the model should connect to a measurable outcome, or it does not belong.

This guide walks through each column and the mistakes that weaken the chain.

Why funders ask for a logic model

A logic model forces discipline that prose hides. When you have to place every activity in a column and draw an arrow to a result, gaps become obvious: an activity with no outcome, or an outcome with no activity feeding it. Reviewers value the model precisely because it exposes whether your program logic actually holds together.

It also serves the reviewer's workflow. A well-built model lets a reader grasp the entire theory of your program in thirty seconds before reading the narrative. For how the model fits the full document, see how a winning proposal comes together, and to start fast, use our logic model template.

Column 1: Inputs

Inputs are the resources you commit to the work: staff, funding, facilities, equipment, partnerships, and volunteer time. List what the program genuinely requires, not everything your organization owns. Inputs should match what appears in your budget, so a reviewer reading both documents sees the same resources.

Column 2: Activities

Activities are the actions you take with those inputs: the tutoring sessions, the trainings, the outreach, the case management. This column is the heart of your methods section in condensed form. Keep each activity concrete and verb-driven so it reads as something you actually do, not a vague intention.

Column 3: Outputs

Outputs are the countable, direct products of your activities. They answer "how much did you do?": 40 workshops delivered, 120 students served, 600 meals distributed. Outputs prove effort and volume, but they are not yet results. The classic error is stopping here and treating activity counts as success.

Column 4: Outcomes

Outcomes are the changes your activities produce in participants: improved reading scores, increased employment, better health, new skills. This is what funders are buying. Outcomes are often split into short-term (changes in knowledge or attitude), medium-term (changes in behavior), and long-term (changes in condition).

The line between outputs and outcomes trips up many applicants. "120 students tutored" is an output. "Reading proficiency rose from 38 to 55 percent" is an outcome. Each outcome in your model should become an indicator in your measurement plan; see how that handoff works in our grant evaluation plan walkthrough.

Column 5: Impact

Impact is the long-term, population-level change you contribute to but rarely cause alone, such as higher community graduation rates or reduced local unemployment. Be modest here. Claiming sole credit for a societal shift undermines credibility; positioning your work as a contributor to it builds trust.

How the columns connect

Read the model left to right as a sentence: with these inputs, we run these activities, which produce these outputs, leading to these outcomes, contributing to this impact. Then read it right to left to test the logic: to achieve this impact, we need these outcomes, which require these outputs, produced by these activities, which need these inputs. If the chain breaks in either direction, fix it before you submit.

The logic model answers what your program does. The deeper question of why those activities should cause those outcomes belongs in a companion document. Pair your model with a clear theory of change when a funder asks for your reasoning, and ground both in tight SMART program objectives.

A single row, read all the way across

Abstract columns become clear when you trace one complete thread from left to right. Take a youth-literacy program. The inputs are two part-time tutors, a donated classroom, a phonics curriculum, and a partnership with two Title I schools. Those feed an activity: 30 weeks of small-group tutoring, three sessions a week. The activity produces an output that is purely countable, 120 students served across 90 tutoring hours. That output drives a short-term outcome, participants' reading fluency scores rising on a fall-to-spring assessment, which leads to a medium-term outcome of more students reading at grade level by year end, contributing to the long-term impact of higher third-grade proficiency across the two schools. Each cell hands off to the next without a gap.

The discipline this exposes is the difference between a model that holds and one that leaks. Notice that the donated classroom in inputs must also appear in the budget as an in-kind contribution, or the two documents contradict each other. Notice that the output (students served) and the outcome (scores rising) are genuinely different claims, one about effort, one about change. And notice that every outcome named is something the evaluation plan can measure with a real instrument, not a hope. When you can narrate one row this cleanly, you can build the rest, and you can spot immediately when a column is carrying a claim the row before it never set up. Keep each thread this tight and the whole page reads as a single, testable argument.

Common logic model mistakes

• Confusing outputs with outcomes. Counting activity instead of change.
• Orphan activities. An activity with no outcome it feeds.
• Overclaiming impact. Taking sole credit for population-level change.
• Mismatched inputs. Resources in the model that never appear in the budget.

Build the chain so every column earns its place, then let the model carry your methods and evaluation. When you want a model reviewers can read at a glance, our grant writing experts build the full inputs-to-impact logic with you.