Impact Measurement: How to Prove Your Work Is Working

A generation ago, most nonprofits reported outputs: "we served 200 youth last year." Today, sophisticated funders (foundations, government agencies, and individual major donors) increasingly expect outcomes: "85% of youth who completed our program reported improved academic performance, and 72% were promoted to the next grade level."

Why the shift? Two forces converged. The sector matured and adopted more of the accountability standards long expected in the for-profit world. Funders started asking not just "did you do the activity?" but "did it work?" At the same time, competition for grant dollars intensified, and outcome evidence became the clearest way for an organization to prove its specific approach deserves funding over the alternatives. Measurement stopped being a "nice to have" add-on and became core to how funding decisions get made.

Before you can measure anything meaningfully, you need a clear theory of change: a logic model that lays out how your activities are supposed to lead to real change. This matters because measurement without a logic model tends to drift toward measuring whatever's easy to count, rather than what actually proves your program works.

Inputs: resources that go in: staff time, space, funding, curriculum. Activities: what the organization actually does: classes, case management, coaching. Outputs: what's directly produced: classes held, participants enrolled, hours of service. Short-term outcomes (0 to 6 months): knowledge gained, skills developed, attitudes changed. Medium-term outcomes (6 to 24 months): behavior changes, condition improvements. Long-term outcomes (2 to 5+ years): ultimate changes: economic self-sufficiency, health improvements, community-level shifts.

Draw this out. Share it with your board. It should directly drive what you choose to measure. If a metric doesn't trace back to a box on this model, it's probably not worth the effort to collect.

Step 1: Define 3 to 5 primary outcome metrics. These should connect directly to your theory of change and matter to funders. Examples: employment programs might track % employed within 90 days, wage rate at placement, and retention at 12 months; education programs might track GPA change, attendance rate, and graduation rate; housing programs might track stability at 90 days, 6 months, and 12 months.

Step 2: Choose your data collection methods. Common approaches include pre/post surveys at intake and exit, follow-up surveys at 3, 6, and 12 months, administrative data (school or employment records, which requires signed consent), structured interviews for qualitative depth, and third-party reporting where available and privacy-appropriate.

Step 3: Build your data systems. Small nonprofits can start with a well-designed spreadsheet. As you grow, purpose-built case management software becomes worth the cost: options range from full-featured platforms like Bonterra or Apricot to Salesforce's Nonprofit Success Pack (powerful but complex to set up) to simple survey tools like Google Forms or Typeform for lighter needs.

Step 4: Assign responsibility. Someone must own data collection: specify who collects what, when, and how. This is the step most organizations skip, and it's the reason most measurement systems quietly die within a year.

Funders want honest data, not perfect data. Report outcomes accurately, including when results are mixed. "Our employment placement rate was 58% against a target of 70%, and here's what we learned about why, and what we're changing" reads as more credible than inflated numbers. Funders who work with many grantees can generally tell the difference, and honesty here builds the kind of trust that gets you renewed.

Context matters. A 58% employment rate for formerly incarcerated individuals with multiple felonies is an extraordinary result. A 58% rate for college graduates in a strong job market would be concerning. Always frame your numbers against what's realistic for your specific population and conditions.

Pair data with stories. One compelling case study (a specific participant, the specific change in their life, told with their permission) makes outcome data human and memorable in a way a spreadsheet alone never will.

This gives you enough to calculate pre/post change for each participant and aggregate it across your whole program: enough to report meaningful outcomes to most funders at your stage, without needing case management software or a research consultant on day one.