Outputs vs. Outcomes vs. SMART Objectives: Fixing the "Vague Goals" Problem in Grant Narratives

"Vague goals" is one of the most common comments a grant reviewer writes on a scoring sheet, and it usually doesn't mean the applicant forgot to set goals at all. It means three related but different things got blurred together: what the project will produce, what will actually change for the people it serves, and the specific, measurable statements that are supposed to connect the two. Those three things have names — outputs, outcomes, and SMART objectives — and freelance grant writers who can keep them straight, section by section, fix one of the most common and most fixable weaknesses reviewers flag.

This guide defines each term the way working grant consultants actually use it, gives a quick test for telling an output from an outcome, and works through a full example so the distinction is concrete rather than academic.

Three different things, in one confusing paragraph

Here's a sentence that mixes all three, the way a first draft sometimes does: "The project will hold 24 workshops, resulting in improved financial literacy, with a goal of reaching 200 low-income adults." A reviewer reading that sentence can't tell what's being promised. Untangled, it's actually three separate claims:

• Objectives are the specific, measurable statements that sit underneath a project's overall goal. Diane H. Leonard, GPC, founder of DH Leonard Consulting, describes them simply: "Objectives are your 'what.'" They're written in SMART format — Specific, Measurable, Achievable, Relevant, Time-bound — and they're what the goals-and-objectives section of a narrative is built around (DH Leonard Consulting, "Grant Objectives vs. Outcomes").
• Outputs are the immediate, countable products of the project's activities — the things you can literally tally from attendance sheets or program logs. Grant consultant Roxanne Jensen, Ed.S., GPC, of Assel Grant Services, describes outputs as "the immediate and direct results of project activities... tangible and quantifiable products that emerge from the project's implementation," tied to the resources and effort put into the project (Assel Grant Services, "Unraveling the Differences Between Goals, Objectives, Outputs, and Outcomes").
• Outcomes are the changes that result from those activities — in knowledge, behavior, or condition — for the people the project serves. DH Leonard's Lead Writer Beth Archer, GPC, puts it this way: outcomes "are what is measurable as a result of what you do," and they indicate whether an objective was actually achieved, where Diane Leonard frames objectives as the "what" and outcomes as the "how well" (DH Leonard Consulting, "Grant Objectives vs. Outcomes").

The fastest test: counting vs. changing

The single most useful rule of thumb, laid out clearly in Grantsights' guide to logic models, is this: "if you can measure it by counting attendance, sessions, or deliverables, it's almost certainly an output, not an outcome. Outcomes require a before-and-after comparison or a threshold achievement that indicates change" (Grantsights, "What Is a Logic Model for Grants?"). "200 people attended the workshops" counts something that happened — that's an output. "70% of attendees scored higher on a financial literacy assessment after the workshop series" measures a change in the people who attended — that's an outcome. The same source also notes that outcomes are usually organized by how soon they show up: short-term outcomes (within about a year) are changes in knowledge or skill; medium-term outcomes (roughly one to three years) are changes in behavior or practice; long-term outcomes (three-plus years) are changes in community-level conditions, which sit closer to a project's broader impact than to anything a single grant period can actually measure.

Where SMART fits — and where sources genuinely disagree

Here's the part worth being honest about instead of smoothing over: not every consultant applies the SMART format to the same box. This article, and the toolkit's own template, treat the SMART statements as the objectives that sit under the goal — the more common convention, and the one DH Leonard Consulting also uses. But Assel Grant Services' own framework writes its outcomes in SMART format instead, defining them as the intermediate results achieved "within one to three years after the project's initiation," tied directly to the objectives above them (Assel Grant Services, "Unraveling the Differences Between Goals, Objectives, Outputs, and Outcomes"). Both are legitimate, published usages from working grant consultants — they simply don't agree on which label carries the SMART formatting.

That inconsistency is a big part of why "vague goals" comments happen in the first place. Diane Leonard notes that grantmakers themselves sometimes use this language in a way that doesn't match any consultant's tidy framework, and her advice — reach out and ask the funder directly when a funder's own instructions are genuinely unclear, rather than guessing — is the most reliable fix available (DH Leonard Consulting, "Grant Objectives vs. Outcomes"). The practical takeaway for a freelance grant writer: check the specific funder's own definitions and application instructions before assuming any one framework, including this one, is universal.

Why the narrative section and the logic model don't use the same boxes

Some of the confusion also comes from the fact that "objectives" and "outcomes" live in two different documents that don't map onto each other one-to-one. Beth Archer at DH Leonard Consulting makes this point directly: "sometimes the confusion comes in because objectives don't have a place in a structure like a logic model" — outcomes are a standard part of a logic model, but objectives, as a labeled box, generally aren't (DH Leonard Consulting, "Grant Objectives vs. Outcomes").

Grantsights' breakdown of the W.K. Kellogg Foundation logic model — the framework most federal agencies reference — lists five components: inputs, activities, outputs, outcomes, and impact. There's no "objectives" box in that chain at all (Grantsights, "What Is a Logic Model for Grants?"). Per the same source, some federal agencies require a logic model as a separate attachment alongside the narrative — SAMHSA for nearly all behavioral health service grants, ACF for most child welfare and family support programs, DOL for workforce development grants, and ED for many education program grants — while NIH and NSF research grants typically don't use the logic-model format at all, relying on different narrative structures instead. A freelance grant writer working across both foundation and federal clients may genuinely need to keep two structures straight for the same project: a narrative goal-and-objectives section, and, for some federal funders, a separate logic model built around outputs, outcomes, and impact. They describe the same project from two angles, and the numbers in one should never contradict the numbers in the other.

A worked example: Cinderbrook Community Pantry

This is a separate fictional case from the one used elsewhere on this site, chosen deliberately so the two examples don't overlap — the organization, funder, and figures below are entirely invented for illustration and shouldn't be reused in a real proposal.

The fictional nonprofit: Cinderbrook Community Pantry, which combines food distribution with a nutrition education class series for low-income families, serving about 500 households per year across its service area in partnership with two regional food banks. The fictional funder: the Fenmoor Foundation, whose stated priorities (invented for this example) include "supporting community-based approaches to food security and nutrition education."

Goal: Improve food security and nutrition knowledge among low-income families served by Cinderbrook Community Pantry.

Objectives (SMART, each labeled as an unconfirmed target until program staff sign off):

• By the end of the grant year, 60% of families completing the nutrition education series will demonstrate improved nutrition knowledge, as measured by a pre/post assessment. (PROPOSED TARGET — CONFIRM WITH PROGRAM STAFF)
• The program will distribute food to at least 500 households across its service area during the grant year. (PROPOSED TARGET — CONFIRM WITH PROGRAM STAFF)
• By the end of the grant year, at least 40% of families completing the nutrition series will report reduced food insecurity, as measured by the USDA six-item food security survey module administered before and after the series. (PROPOSED TARGET — CONFIRM WITH PROGRAM STAFF)

Outputs — what the countable records will show: the number of food distribution events held over the grant year; the number of households served at each distribution; the number of nutrition education classes held and attendance at each. Every one of these is a tally staff can pull straight from a distribution log or a class sign-in sheet — no before-and-after comparison required.

Outcomes — what those records can't show by themselves: whether families' nutrition knowledge actually improved (the first objective), and whether their food security actually improved (the third objective). Both require comparing a "before" state to an "after" state — an assessment score, or a survey response — not just a count of distributions held or households served. Notice that the second objective ("distribute food to at least 500 households") is really phrased as an output dressed up as an objective: it's a count, not a change. That's not a mistake unique to this example — a project can legitimately have a reach or capacity objective even though, strictly speaking, it measures the same kind of thing an output measures. The distinction matters most for the objectives that are supposed to demonstrate impact, not capacity.

The general mechanics of an evaluation plan — what to measure, how, how often, and who's responsible for it — are covered in more detail in Grant Narrative Structure Explained, which walks through a different fictional program from a different field. For Cinderbrook specifically, the plan that actually connects the outputs above to the outcomes the objectives promise names three measurement methods: the pre/post nutrition knowledge assessment measures the first objective, distribution and enrollment tracking measures the second, and the USDA six-item food security module measures the third. If any of those three measurement methods were missing, the objective it's supposed to support would have no way of being proven true or false by the time the final report is due.

A short checklist before you submit

• Does every objective actually measure a change, not just a count — or if it measures a count (like enrollment), is that an intentional, honest choice rather than an oversight?
• Is every output something the organization can pull from records it already keeps, with no invented or projected numbers?
• Is every numeric target in an objective labeled as unconfirmed until someone who runs the program signs off on it?
• Does the evaluation plan name a specific measurement method for each outcome — not just a restated hope that it will happen?
• If this funder's own guidelines define "objective," "output," or "outcome" differently than the working definitions here, does the narrative follow the funder's definition instead?