Measuring your nonprofit starts with a simple question: What are you trying to achieve?
“Explain to me how what you’re doing is going to achieve this goal,” says Marc Epstein, a professor at Rice University and author of the book Measuring and Improving Social Impacts. “And let’s be really clear about what this goal is.”
That may sound simple, but Mr. Epstein says many nonprofits don’t have a clear idea of what they’re trying to achieve.
Your organization wants to make students better-prepared for the 21st-century workforce? Great. But what does better-prepared mean? Do you want to raise their math scores? Boost their rates of acceptance to college?
Once your nonprofit has a specific goal, it must have a hypothesis about how its program or service -- often called an “intervention” -- will produce the desired change. This is your “logic model” or “theory of change.”
“There ought to be some sort of causal relationships between your inputs and what you expect to get out of it,” says Mr. Epstein.
Again, this may sound basic, but it is the foundation for any system of measurement. You need to know what you want to achieve and how you plan to achieve it -- and be specific.
Understanding Logic Models as Measurement Tools
To help with impact evaluation and to more clearly understand how process affects impact, it’s helpful to create a logic model.
Also called a “program theory,” a logic model is simply a flowchart that maps the sequence of events intended to effect a change. It shows the relationships among a nonprofit’s resources, how it uses them in programs, and the results of those programs.
According to the W. K. Kellogg Foundation, the components of a logic model are:
-
Inputs/Resources: “The human, financial, organizational, and community resources a program has available to direct toward doing the work.” For example, if your nonprofit runs a job-training program, the inputs are the money available to buy supplies, the facility available to host the events, and the staff members or volunteers available to work.
-
Activities: “The processes, tools, events, technology, and actions that are an intentional part of the program implementation.” For the job-training program, activities may include monthly career-skills workshops, one-on-one resume critiquing sessions, and a twice-annual career fair with local companies.
-
Outputs: “The direct products of program activities and may include types, levels, and targets of services to be delivered by the program.” The number of people who complete the job-training program is one output.
-
Outcomes: “Specific changes in program participants’ behavior, knowledge, skills, status, and level of functioning.” The number of participants who found jobs explicitly because of completing the program is an outcome.
-
Impact: “Fundamental intended or unintended change occurring in organizations, communities, or systems as a result of program activities.” A change in the local community’s unemployment rate could be an impact of the program.
Some logic models include process evaluations between the activities and outputs stages to measure how well services were provided and activities carried out.
It’s important to measure and report as far down the logic-model chain as possible in order to fully understand a nonprofit’s effect, says Jacob Harold, chief executive of GuideStar, a nonprofit information service.
The most important distinction to draw in the model is between outputs and outcomes, says Mr. Harold.
Some nonprofits conflate the two, only reporting the outputs of their work without proving that it led to any real change. For example, a job-training program that reports only its outputs -- 100 people attended a workshop -- does not present a compelling case that its programs actually help people find employment, only that it successfully attracted participants.
According to Mr. Harold, if a nonprofit is unable to fully measure outcomes, it should still present a hypothesis or theory of change about why its activities lead to certain results that is backed by as much evidence as possible.