If an evaluation section has ever come back marked "not measurable", the cause was probably not the evaluation. It was the objectives. Outcome sections often fail because they describe what the organization will do, not what will change for the people it serves.
This article gives you the vocabulary, shows how to tell an output from an outcome, and walks through the four things every measurable objective needs. You'll also see how to use a funder's own performance measures and a simple logic model so your plan reads the way the funder expects.
Learn the four terms and keep them apart
These words get used loosely, which is why proposals blur them. A working set of definitions:
- Goal: the broad change you want. It's the destination, and it usually takes longer than the grant.
- Objective: a specific, measurable step toward the goal, within the grant period.
- Output: what you do or produce. Sessions held, people served, materials distributed.
- Outcome: what changes for people because of what you did. Knowledge, behavior, condition or status.
Funders sometimes define these terms themselves, and the notice may use them differently from the way you do. Use the notice's definitions where it gives them.
Outputs count activity; outcomes count change
An output is easy to count and easy to promise. An outcome is harder, because it depends on whether the work made a difference. Reviewers know the gap, and outcomes are where they look hardest.
Here are two hypothetical objectives from a youth mentoring program:
| Type | Example |
|---|---|
| Output | Serve 200 youth in the mentoring program. |
| Outcome | Increase school attendance among the 200 participating youth by 15% by the end of the second school year. |
The first tells the reviewer how big the program is. The second tells them what the program is for. Funders usually want both, since outputs show you can deliver at scale. But when scores are tallied, the outcome carries the weight.
A check you can run on any objective: ask "so what happened to the people?" If the sentence can't answer, it's an output.
Give every objective four parts
A measurable objective has a baseline, a target, a time frame and a data source. Leave one out and a reviewer has to guess.
- Baseline: where things stand now. Without it, a target means nothing.
- Target: where you expect them to be.
- Time frame: by when, inside the period of performance. Under 2 CFR 200.1, the period of performance is the time between the start and end date of the federal award.
- Data source: how you'll know. A school attendance report, a pre and post survey, case records.
Say a hypothetical food pantry wants to reduce the number of households that run out of food before month's end. Here is a version without the four parts, and one with them:
Weaker: We will improve food security for the families we serve.
Stronger: By the end of year one, the share of enrolled households who report running out of food before the end of the month will fall from 62% (intake survey, January) to 45% (follow-up survey, December).
The second can be scored. It can also fail, which is why it is credible. An objective that can't fail isn't measuring anything.
If you don't have a baseline yet, say so and say how you'll collect it. "Baseline to be established in the first 90 days through the intake survey" is a sound answer. A made-up baseline is not.
Be careful about targets
Pick targets you can defend. A target pulled from the air raises questions; one tied to your past results, a published benchmark or the funder's own measure does not. Say a mentoring program saw attendance rise 12% in a previous cohort. A 15% target for a slightly longer program is a reasonable stretch, and you can say why.
Use the funder's own measures where they exist
Many federal programs set their own performance measures. Under 2 CFR 200.301, a federal agency should clearly communicate the outcomes, indicators, targets, baseline data or data collections it expects a recipient to measure and report. The same section says agencies measure performance to show achievement of program goals and objectives and to improve program outcomes.
For you, that means two things. First, check the notice for required measures before you invent your own. When the funder names a measure, use its exact wording and definition so your numbers can be added to everyone else's. Second, if the notice says reporting will use specific indicators, build your objectives around them and don't make the funder translate. See also how to read a funding notice, which covers where these requirements tend to sit.
Draw a simple logic model
A logic model is a picture of the chain from resources to results. A common structure is:
- Inputs: staff, money, partners, space.
- Activities: what you do with them.
- Outputs: what those activities produce.
- Short-term outcomes: early changes, such as knowledge or skills.
- Long-term outcomes: lasting changes in behavior or condition.
The U.S. Department of Education defines a logic model in its regulations (34 CFR 77.1) as a framework that identifies the key components of a project and describes the relationships among those components and the outcomes they are meant to produce. That is a good test for your own: each outcome should trace back to a component, and each component should lead somewhere.
If the funder supplies a logic model template or its own performance measures, use them rather than building your own. If it doesn't, a one-page table is enough. It will also expose weak spots. If you find an outcome with no activity feeding it, or an activity that leads to no outcome, you've found something to fix before a reviewer finds it for you.
A rewrite exercise
Take one objective from an old proposal and ask four questions:
- What is the baseline?
- What is the target?
- By when?
- How will we know?
Then ask the fifth: is this an output or an outcome? If it's an output, keep it, and add the outcome it is supposed to produce.
Once your objectives are clear, the evaluation plan becomes much easier to write. It's mostly a matter of saying how you'll collect the data you just named. See evaluation plans for people who aren't evaluators and metrics that matter for the next steps. The need statement is also a useful source of baselines, since it already holds the figures that describe where things stand today.
Your next step
Take one objective from a past proposal and rewrite it with a baseline, a target, a date and a data source. If you can't fill one of the four, you've found the question to answer before your next application.
SoliContraX can help you draft outcomes from your own program materials and check them against what the notice asked for, so you aren't starting from a blank page. See how to get started.