How to build a logic model: a practical guide with an example
A logic model explains how a program is expected to work. It connects the resources you invest, the activities you deliver, and the changes you hope to support. Use it to plan with your team, explain a grant proposal, or decide what to evaluate.
An interactive logic model keeps those relationships editable as your understanding changes. In dylomo, you can examine the same program in Traditional, Flow, and Dynamic layouts, then prepare it for discussion or export.
Example: a youth mentorship program
Suppose a community program wants young people to have dependable adult support. This illustrative pathway is a starting point for discussion, not evidence that mentoring will produce these results in every setting.
- Resources / inputs
- A coordinator, trained volunteer mentors, meeting space, and transport support.
- Activities
- Match young people with mentors, hold weekly sessions, and practise setting personal goals.
- Outputs
- Mentoring matches made, sessions delivered, and young people participating regularly.
- Short-term outcomes
- Participants can name a trusted adult and describe a realistic personal goal.
- Intermediate outcomes
- Participants practise asking for support and follow through on agreed goals.
- Long-term outcomes
- Stronger connection to school and community, alongside support from families and other services.
The pathway depends on assumptions: mentors can attend consistently, young people find the sessions useful, and safe meeting spaces are available. Transport, school schedules, and changes in other services are external factors to discuss.
Outputs and outcomes answer different questions. “Twenty sessions delivered” describes what the program produced. “Participants feel more confident asking for support” describes an intended change. Neither statement alone proves the program caused that change.
Build your own logic model in five steps
- Agree on the need and the people you want to reach. Write a short description of the situation with program staff, participants, and partners. Decide which part of the need this program can realistically address.
- Describe the intended changes. State who or what should change, in what way, and over what timeframe. Work back from those outcomes to the activities and resources they require.
- Separate delivery from results. List activities, their immediate outputs, and the expected outcomes separately. In the mentoring example, holding a session is an activity; participation is an output; using a new skill is an outcome.
- Make the relationships explicit. Connect each activity to the changes it is expected to support. Describe why the relationship is plausible, what evidence informs it, and which assumptions still need testing.
- Choose what to learn and measure. For a priority outcome, agree on an evaluation question, a useful indicator, a data source, and when to review it. Keep that evaluation plan alongside the model.
Review the pathway before sharing it
- Does every activity have a clear purpose in the pathway?
- Have you described a plausible step between immediate outputs and long-term outcomes?
- Are assumptions visible, including who may face barriers to participating?
- Which outcomes depend on other programs, partners, or wider conditions?
- Could a participant or colleague explain the model back to you?
A theory of change can provide a fuller explanation of why and under what conditions change is expected. A logic model makes a selected pathway easier to examine together. Keep the explanation and the model consistent as the program develops.
Use dylomo to develop the model with your team
Start with a blank model or a built-in methodology. Add your program components, describe their relationships, and compare visual layouts to see which communicates the pathway clearly. Free supports core model creation and exports. Model owners need active Pro to invite collaborators or publish public links; invited people can use Free accounts.
Further reading
For program-description guidance, see the CDC’s evaluation framework: describe the program. The University of Wisconsin Extension’s logic model glossary explains common evaluation terms. The mentorship example above was written for this guide and does not describe a measured program.