A federal grant logic model is more than a required attachment. It is a visual explanation of how your organization will use its resources, carry out program activities, measure results, and achieve intended outcomes. Most nonprofits treat the logic model as a form to be completed. Federal reviewers treat it as a window into program thinking. The difference in how those two groups approach the same document explains a significant portion of applications that lose points on program design — not because the program is weak, but because the logic model does not demonstrate that the applicant understands what their program is actually doing.
A logic model is a visual representation of how a program is supposed to work: what resources go in, what activities are carried out, what products or services result, what changes occur in participants, and what broader impact is intended. The Office of Juvenile Justice and Delinquency Prevention, which funds programs parallel in structure to OVW and OVC grants, describes logic models as the conceptual backbone connecting program design to performance measurement. That framing is accurate. A logic model is not decoration — it is a claim about cause and effect.
The Five Components and What They Actually Represent
Inputs are the resources the program brings to bear: funding, staff, partner organizations, physical space, existing client relationships, and organizational expertise. Inputs are not aspirational. If you list “community trust” as an input, you should be able to describe specifically how that trust was built and how it translates into program access.
Activities are what the program does with those inputs. For a victim services program, activities might include crisis intervention hotline services, safety planning sessions, court advocacy, transitional housing placement, and trauma-informed case management. Activities should be specific enough that a reviewer can imagine staff carrying them out.
Outputs are the countable products of those activities: number of clients served, hotline calls answered, court accompaniments provided, housing placements made. Outputs are measurable but do not, by themselves, indicate that anything changed for participants. This is the most common logic model failure — listing outcomes that are actually outputs. “200 clients received case management services” is an output. It describes what you did, not what happened as a result.
Outcomes are the changes that occur in participants because of the activities: increased safety, reduced isolation, improved access to legal resources, stabilized housing, and restored sense of agency. Short-term outcomes are the immediate changes after program contact. Medium-term outcomes develop over months. Long-term outcomes — sustained safety, economic stability — may extend beyond the grant period.
Impact is the broader community-level change that results if the program succeeds at scale: reduced rates of domestic violence-related homicide, decreased emergency shelter utilization as housing-first models take hold, stronger coordinated community response infrastructure.
Why the Logic Model Must Match the Narrative and Budget Exactly
Federal reviewers read the logic model alongside the proposal narrative and the budget. They are looking for internal consistency. If the narrative describes a legal advocacy component but no legal services appear in the logic model, the gap is visible. If the budget funds a full-time housing navigator but outcomes related to housing stability do not appear in the logic model, reviewers will wonder whether the position was added opportunistically rather than as a designed program element.
The logic model functions as a summary of everything else in the application. Reviewers use it to check whether the applicant has a coherent theory of change — whether they can explain not just what they will do, but why doing it will produce the results they claim. A strong logic model makes the narrative easier to evaluate because it provides a framework for everything that follows.
Common Logic Model Failures
The failures we see most often fall into three categories.
The first is vagueness in activities. “Provide services to survivors” is not an activity description. “Conduct weekly individual safety planning sessions with clients, using the Danger Assessment tool to guide lethality screening” is an activity description.
The second is outcomes that are actually outputs. Organizations list “clients will receive 12 weeks of counseling” as an outcome. That is a service delivery count. An outcome would be “clients will demonstrate reduced trauma symptoms, measured by validated screening tools at intake and exit.”
The third is a problem statement that does not connect to the logic model. If the needs assessment establishes that underserved rural survivors face transportation barriers as the primary obstacle to safety, the logic model should show an input (transportation resources or partnerships), an activity (mobile outreach or transportation assistance), and an output and outcome that flow from that. If the needs assessment and the logic model appear to describe different programs, reviewers will notice.
What a Strong Logic Model Looks Like for a Victim Services Program
Consider a hypothetical transitional housing program for domestic violence survivors. A well-constructed logic model might look like this:
Inputs: OVW grant funds; licensed social work staff; partnerships with legal aid, housing authority, and employment services; 8 transitional housing units; trauma-informed training completed by all direct service staff.
Activities: Individualized safety planning at intake; weekly case management meetings; financial literacy workshops; legal advocacy and court accompaniment; warm referrals to partner employment and childcare services; monthly partner coordination meetings.
Outputs: 40 survivors served annually; 480 case management sessions; 24 financial literacy workshops; 75 court accompaniments.
Short-term outcomes: Clients report increased safety knowledge and reduced immediate danger (measured by Danger Assessment at intake and 90-day follow-up); clients identify at least two community support resources.
Medium-term outcomes: 70% of clients maintain stable housing for 12 months post-exit; 60% achieve economic independence milestones (employment, benefits access, or completion of vocational training).
Impact: Reduced shelter recidivism and increased long-term housing stability among domestic violence survivors in the service region.
This is the level of specificity that distinguishes a logic model that earns points from one that satisfies a checkbox. The connections between each column should be legible to someone who knows nothing about your organization but understands program design.
If your organization wants to build grant applications that hold together at this level of rigor, start with our grant readiness framework — because a logic model that works begins with program clarity that exists before you open the NOFO.
