Asked by a foundation "what results are you planning," an organization confidently answers: "we'll run twelve workshops." The answer feels concrete and convincing to them, there it is, twelve, not two.
The foundation reads that answer differently: the organization doesn't distinguish its own activity from change in people's lives. A workshop isn't a result, it's what you do. A result is what changed for participants afterward. The difference between these two things is this lesson's whole point.
A logic model is a chain linking what you put in to what ultimately changes: resources, activities, outputs, outcomes, impact. The key difference within it, between outputs (what you produced) and outcomes (what changed), is the difference between "we worked" and "we changed something," and that's exactly what a funder is looking for.
Resources are what you put in: money, people, time, space. Activities are what you do with those resources: running sessions, consulting, distributing materials. Outputs are the measurable volume of what got done: twelve workshops, sixty participants, three hundred kits distributed.
Outcomes are what changed as a result: forty participants applied the skill they learned, twenty-five saw a specific measurable indicator improve. Impact is the long-term shift all of this leads toward. Most organizations confidently describe resources, activities, and outputs, and stumble exactly at outcomes, because outcomes require admitting that the volume of work alone proves nothing yet.
An honest small metric beats an imaginary big one. "We'll change a generation's life" sounds impressive but can't be verified, so it's worth nothing. "The share of participants completing the program grew from forty to seventy percent" sounds more modest, but it's real proof a foundation believes.
A metric gets designed before the program launches, not remembered in a panic before a report is due. A simple before-and-after survey, the share who finished, a short skill test, all of this needs to be thought through in advance, so there's something to measure. A metric invented after the fact almost always turns out either unmeasurable or unconvincing, because nobody collected the data for it in time.
The most valuable thing about a logic model isn't that it's needed for the application, it's that it exposes gaps in the program itself. If your stated outcomes don't logically follow from your activities, the program is decorative: you're doing something, but you don't understand what change it leads to.
Many organizations first truly understand their own program exactly when they first draw its model and discover a gap between what they do and what they want to change. That's why it's worth building a logic model even without a grant application in the picture: it's a diagnostic of program health, and an application is just one of its uses.
Below is a builder that assembles your main program's logic model step by step, checking at each link whether it's really an output, not already an outcome.
Your main program's logic model. This is one of the most reusable documents in your whole Binder: it goes into grant applications, donor reports, the annual report, and explanations of the program to new board members. Built well once, it saves dozens of hours down the line.
What's the exact difference between an output and an outcome?
An output is the volume of what you did (twelve sessions, sixty participants). An outcome is the change in people as a result (a skill was learned, an indicator improved). The first is about you, the second is about them.
What if outcomes are hard to measure in my field?
There's almost always at least a simple, honest metric available: a survey, participants' self-assessment, completion rate. A small measurable metric beats a large unmeasurable one.
Do I have to build a logic model if a foundation doesn't explicitly ask for one?
Not formally, but it exposes weak spots in the program regardless of any application. It's worth building at least once, even just for your own internal understanding.
Can I use one logic model for different foundations?
The core of one program stays the same, but the emphasis should adjust to each specific foundation's priorities. The chain itself stays honest and unchanged.
A logic model shows how your actions turn into change. But every link in that chain has a cost, and the next reader of your application will most likely look at exactly that. The next lesson is about the grant budget, the document that either confirms your model with numbers or destroys trust in it with one mismatch.