Pro free for 30 days.No card.Start free →

Measuring impact and outcomes: a practical nonprofit guide to outputs, outcomes and indicators

Counting what you do is not the same as showing what changed. This guide explains outputs, outcomes and impact, how to build a logic model, choose one to three indicators, collect data lightly and report results honestly to donors and funders.

By WorkBOS Content Desk7 min read
Magnifying glass over printed tables and business documents (Photo by ds_30 on Pixabay)
Photo: ds_30 / Pixabay

Ask a nonprofit leader whether measuring impact matters and the answer is almost always yes. Ask whether they feel good at it and the room goes quiet. Candid, the nonprofit information and training group, has cited a survey in which 76% of nonprofit leaders called impact measurement a top priority, but only 29% felt very effective at demonstrating outcomes. The gap is not about caring. It is about time, tools and knowing where to start.

This guide to measuring impact and outcomes is written for small nonprofits with limited staff and no evaluator. You will learn the difference between outputs and outcomes, how to build a simple logic model, how to choose a few useful indicators, how to collect data without burdening people, and how to report results honestly. It is general guidance, not evaluation or legal advice.

Outputs, outcomes and impact: three words that get mixed up

Clear terms prevent muddled reports.

  • Outputs are what you do and count: meals served, people trained, hours of counseling, workshops held.
  • Outcomes are the changes that follow for the people or places you serve: improved confidence, stable housing, better grades, healthier food access.
  • Impact is the longer-term, broader change in the community, often influenced by many factors beyond your own work.

A nonprofit data guide from Nonprofit Hub illustrates the difference: a food pantry counting 500 community members served and 1,500 cans distributed has outputs, while a conclusion about improved food security among low-income households is an outcome. A career seminar can show an outcome by comparing participants' before and after self-scores on interview confidence. Outputs are still essential, since they are the foundation outcome work is built on.

Start with a simple logic model

A logic model is a one-page picture of how your work is supposed to lead to change. The CDC's widely used evaluation framework describes it as a graphic depiction connecting activities to intended short-, medium- and long-term effects.

The five columns

| Column | Question | Example: tutoring program | |---|---|---| | Inputs | What do we invest? | Volunteers, space, books, budget | | Activities | What do we do? | Twice-weekly tutoring sessions | | Outputs | What do we count? | 60 students, 400 sessions | | Outcomes | What changes? | Better reading scores, more confidence | | Impact | What long-term change do we contribute to? | More students graduating on time |

Draw it with staff, volunteers and, ideally, people you serve. Disagreements that surface here are valuable; they reveal different beliefs about what the program is for.

Follow a lightweight evaluation cycle

The CDC framework summarized by CASRAI has six steps. The idea holds for small organizations if you scale it down.

  1. Engage stakeholders: involve those who run the program, those it serves and those who will use the findings.
  2. Describe the program: write down its need, activities, resources and logic model.
  3. Focus the evaluation: decide the purpose, the users and a few key questions, since no evaluation can cover everything.
  4. Gather credible evidence: choose indicators, sources and collection methods that fit your questions and your limits.
  5. Justify conclusions: set standards for judging performance, interpret the evidence and state a judgment.
  6. Use and share lessons: plan from the start for how findings will lead to action.

The guide describes these as a cycle you can return to, and it adds that the framework applies four standards across all steps: utility, feasibility, propriety and accuracy. The CDC revised the framework in 2024 with updated labels, so check current CDC materials if you want the newer wording. The big lesson is to ask for only the evidence you will actually use.

Choose a few good indicators

An indicator is a specific measure that shows progress. Resist the urge to track everything.

Rules of thumb

  • Start from your goal. Nonprofit Hub advises clarifying long-term goals first, because they decide which outcomes to track. A college-prep program might follow acceptance rates and how many participants can afford their chosen school.
  • Pick one to three per program. More than that and nobody reads them.
  • Mix quantity and story. Candid recommends combining numbers with narratives that explain attitudes and experience.
  • Use what already exists. School records, intake forms and attendance logs may already hold usable data.
  • Define each one in writing: what is counted, who is counted, when and by whom.

Examples

| Program | Output | Outcome indicator | |---|---|---| | Job training | Participants completed | Share employed 90 days after finishing | | Food pantry | Households served | Share of clients reporting fewer skipped meals | | Youth mentoring | Matches made | Change in school attendance or self-reported confidence | | Housing counseling | Clients counseled | Clients avoiding eviction after six months |

Collect data without wearing everyone out

Make collection part of normal work

Capture information at the moment of service: intake forms, check-in sheets, short surveys after sessions. Nonprofit Hub warns that paper forms and messy spreadsheets can slow you down and skew results, and suggests centralized tracking, such as a case management or database tool, so data is ready for analysis.

Keep surveys short and respectful

Ask five to ten questions, use plain language, offer different ways to answer and explain how responses will be used. Do not collect personal information you do not need. If you work with children, health or other sensitive information, check consent and privacy requirements first, and see our guide on AI and customer data privacy if you use AI tools on the data.

Follow up with a few people

Interviews and short stories add the explanation behind the numbers. Ask permission before quoting anyone, and protect identities.

Look at results and report honestly

Review on a schedule

Candid distinguishes management, which monitors processes and is often reviewed monthly, from strategy, which looks at longer-term effects and is reviewed twice a year or annually. Nonprofit Hub cautions against checking outputs only at the end, since regular reviews leave room to adjust. A quarterly review of your indicators, with staff and a board member, is a good start.

Interpret carefully

  • Compare to a baseline, such as before and after, or last year and this year.
  • Be cautious about claiming that your program caused a change when other factors are involved. Say "participants improved" rather than "we fixed".
  • Report what did not work, and what you will change.
  • Do not rely solely on impact-per-dollar figures, which Candid says do not prove effectiveness by themselves.

Tell the story to different audiences

Donors want a clear headline and a human example. Funders want the numbers and the method. Your board wants trends and decisions. Pair every chart with a plain-language sentence, since raw data without explanation can confuse. Funder-specific reports are covered in our grant tracking and reporting guide, and thanking supporters with results is covered in donor management.

Common mistakes

  • Counting only activities. Busy is not the same as effective.
  • Too many indicators. A long list produces data nobody uses.
  • Collecting data you never analyze.
  • No baseline. You cannot show change without a starting point.
  • Overclaiming. Taking sole credit for community-wide change.
  • Ignoring the people served. They often know best what matters.
  • Waiting for a perfect system. Start with one program and a few measures.

How WorkBOS can help

WorkBOS includes surveys and forms for collecting feedback, projects and tasks for tracking program activities, CRM records you can adapt for participants, and drives for storing reports. You can build simple summaries from the data, and AI agents can suggest draft write-ups, always acting only with human approval so someone checks the numbers first. It is a general business tool rather than an evaluation platform, so it suits lightweight tracking more than formal research. See pricing, the docs and the Academy. For broader record work, read running volunteers well as well.

Frequently asked questions

Q: What is the difference between outputs and outcomes? A: Outputs count what you do, such as people served. Outcomes describe what changed for them, such as improved skills or stability.

Q: How many indicators should we track? A: One to three per program is plenty to start. Add more only when you are using the ones you have.

Q: Do we need an outside evaluator? A: Not to begin. A simple logic model, a few indicators and honest review go a long way. Consider an evaluator for large funders, formal research or complex programs.

Q: How often should we review results? A: Monitor monthly or quarterly, and do a deeper review once or twice a year before planning and budgeting.

Q: Can we share results if they are disappointing? A: Yes, and you should. Explain what you learned and what you will change. Funders generally respect candor.

Sources are linked in the text above: CASRAI (CDC evaluation framework guide), Nonprofit Hub (Candid guidance is also referenced).

Run it all from one place

Projects, clients, finance, people and AI employees in one system. Start free, no card needed.

Start free →