Small organizations often write the evaluation section last, in a hurry, with no data person on staff. The result is a page of promises: surveys, interviews, a dashboard, an outside evaluator. Nobody on the team is sure they could deliver half of it.
You don't need an evaluation degree to write a plan a reviewer trusts. A credible plan answers four questions: what you will measure, how you will collect it, who will do it and what you will do with the answers. This article walks through each, including how to tell process from outcome evaluation and when an outside evaluator is worth the cost.
Know the two kinds of evaluation
Evaluation comes in several types, but two cover most small-grant plans. The CDC's evaluation guide describes them this way:
- Process (or implementation) evaluation assesses how well the program is being implemented as planned. In plain terms: did we do what we said we would do?
- Outcome evaluation measures how well a program has achieved its intended outcomes. In plain terms: did it make a difference?
The same CDC page adds a caution worth repeating in your plan: an outcome evaluation can show whether outcomes were achieved, but it cannot determine what caused them. Don't claim more than your design can support. "Participants' reading scores rose by an average of eight points" is something you can show. "Our program raised the scores" is a claim that needs a comparison group.
Most proposals need both. Process data tells you whether a weak result came from a weak program or from a program that never ran as designed. Outcome data tells you whether the effort paid off. Tie your outcome measures to the objectives in your proposal, and if those are still fuzzy, start with goals, objectives and outcomes.
Start from data you already collect
The best data collection plan is one your staff can actually carry out. Before you design anything new, list what you already gather:
- Attendance or enrollment records
- Intake forms and case notes
- Surveys you already give at the end of a program
- Referral and follow-up logs
- Records from partners, such as school attendance or employment data, if you have a data-sharing agreement
Then match each to something you said you would measure. Say a hypothetical after-school tutoring program has these objectives and sources:
| What we'll measure | How | Who | When |
|---|---|---|---|
| Sessions delivered vs. planned (process) | Tutor sign-in sheets | Program coordinator | Monthly |
| Students attending at least 80% of sessions (process) | Attendance log | Program coordinator | Each semester |
| Change in reading level (outcome) | Fall and spring reading assessment the school already gives | Coordinator, with school records | Fall and spring |
| Student and family satisfaction (outcome) | Short end-of-year survey | Coordinator | May |
Every row has a source and an owner. Nothing requires a new database.
If a measure calls for a validated tool, such as a standard survey or screening instrument, name it and say it's well established. Reviewers like seeing that you didn't invent a scale. Check the notice, too, because many funders require particular measures or reporting formats. Under 2 CFR 200.301, a federal agency should communicate in the award any expected outcomes, indicators, targets, baseline data or data collections the recipient is responsible for measuring and reporting, and it should specify any requirement to participate in federally funded evaluations.
Don't promise what you can't staff
A common weakness in small-organization evaluation plans is a plan that outruns the people. Interviews with every participant, quarterly focus groups and a pre/post/follow-up survey all sound strong, but they take hours.
A practical test: for each data collection activity, name the person who does it and estimate the hours. If the hours don't fit anyone's workload, cut the activity or budget for help. A short plan you complete beats a long plan you don't, and reviewers know it.
Decide between internal and external evaluation
Some funders require an external evaluator. Some leave it to you. Some award points when you have one. Read the notice before deciding.
Even when it's optional, an outside evaluator may be worth it if you need credible outcome evidence, if the program is new or you plan to scale it, or if you don't have anyone who can design a measurement approach. A local university, a research center or an independent evaluator can all fill the role. The article on partnering with universities and evaluators covers how to find and work with them.
Budget for the work. Under 2 CFR 200.459, costs of professional and consultant services by people who aren't your employees are allowable when reasonable in relation to the services rendered and not contingent upon recovery of the costs from the federal government. The rule lists factors that bear on allowability, including the nature and scope of the service in relation to the service required and the qualifications of the person providing it. Describe the evaluator's scope in the budget narrative, as the budget justification line by line explains. Confirm cost rules for your specific award with your grants officer.
Say what the evaluator does, how often they report and who owns the data. If you're using an internal evaluation, name the person responsible and explain how you will keep the work honest, for example by having someone outside the program review the findings.
Say what you'll do with the results
An evaluation that ends in a report nobody reads isn't doing its job. The CDC's evaluation framework ends its steps with "act on findings", and says insights should be used throughout. Tell the reviewer how you will do the same:
- Improve the program: a mid-year review where staff look at process data and adjust.
- Report to the funder: what you'll submit and when.
- Share with the people you serve: a short summary of what you learned.
- Inform the next proposal: results that go into your records for future applications.
One sentence on each is enough. "We will review process data each quarter and adjust recruitment if enrollment falls below 90% of target" is specific. "We will use results for continuous improvement" is not.
Your next step
List every piece of data your organization already collects, who collects it and how often. That list is the start of your next evaluation plan, and it will show you which of your objectives you can already measure.
SoliContraX keeps your materials, including evaluation tools and past results, in one place, then helps you draft an evaluation section and shows what is still missing. Get started with SoliContraX.