Writing the Proposal

A grant proposal wins by discharging six burdens of proof in order: the problem is real, the solution follows from it, the applicant can execute, success will be measured, the money is derived, and the work survives the grant. Reviewers score whether each burden was met.

Key takeaways

• Every proposal section exists to discharge one burden of proof.
• Section names change by funder; the burdens do not.
• Reviewers catch inconsistency between sections before weakness inside one.
• Compliance failures end applications before any score is assigned.
• Two burdens — money and measurement — have their own hubs.

What sections does a grant proposal contain?

Nearly every grant proposal contains the same eight components under different labels: a summary, a statement of need, goals and objectives, a project design, an organizational capacity section, an evaluation plan, a sustainability plan, and a budget with a narrative. The reference on the anatomy of a grant proposal maps those components across the vocabularies that the National Institutes of Health, the National Science Foundation, federal program offices, and private foundations each use.

Federal announcements converge on that set because the set is prescribed. Appendix I to 2 CFR Part 200 requires every notice of funding opportunity to state its application contents, its review criteria, and its submission rules (Appendix I to 2 CFR Part 200). The Department of Education publishes a menu of general selection criteria — need, project design, personnel, resources, management plan, and evaluation — that program offices draw from when writing a competition (34 CFR 75.210).

Foundations compress the same material rather than dropping it. A two-page foundation narrative still asks for need, plan, capacity, measurement, and money; the sections have merged. Mistranslating between vocabularies is where a grant proposal loses points for content it actually supplied.

How does a grant proposal prove the problem is real?

A grant proposal proves the problem is real by documenting a gap — two data points and the distance between them — for a named population in a named geography. Writing a statement of need covers the construction in detail, including the rule that decides most drafts: the need belongs to the community, never to the applicant organization.

The single most common defect is circularity. A statement that the absence of the applicant’s program is the problem describes a want, not a need. The working test is whether the problem would still be describable if the organization vanished.

Comparison is not decoration in a grant proposal; it is the scored criterion. Federal reviewers are told to weigh “the data presented (including a comparison to local, State, regional, national, or international data) that demonstrates the issue, challenge, or opportunity to be addressed by the proposed project” (34 CFR 75.210). A prevalence figure with no benchmark establishes that a condition exists somewhere. It does not establish a gap, and a gap is what the section is asked to prove.

What is the difference between a goal and an objective?

A goal states the direction a grant project moves in; an objective states a specific, measurable change the project commits to producing. Goals, objectives, and activities works through the full chain — goal, objective, activity, output, outcome, impact — and shows where each link is checked.

Funders buy objectives. A goal is unobservable by design, so a grant proposal whose objectives merely restate the goal has committed to nothing a reviewer can score or a program officer can later verify. Objectives carry a population, a magnitude, a baseline, and a date; without a baseline, no reader can judge whether a target is ambitious or trivial.

The SMART acronym is a useful last-pass format check and a poor generator. It catches an objective missing a number or a deadline, and it says nothing about whether the objective follows from the documented need — which is the property reviewers actually score. Run it after the objective set has been derived from the need and the design, never before. Objectives written at proposal stage also become the reporting obligation after award, which is a reason to choose them carefully rather than competitively.

How does the project design show the plan can work?

The project design section shows the plan can work by specifying activities, sequence, dosage, staffing, and participant flow at a level where a reviewer can check the arithmetic. Writing the project design treats the section as an argument about mechanism rather than a description of intent.

Dosage is where a grant proposal design becomes checkable. A program promising a full grade level of reading gain from one short session a month has an arithmetic problem no prose conceals. Stating frequency, duration, group size, contact hours per participant, and expected attrition converts an aspiration into a claim.

The National Science Foundation writes the standard into its merit review criteria as a pair of questions: “Is the plan for carrying out the proposed activities well-reasoned, well-organized, and based on a sound rationale? Does the plan incorporate a mechanism to assess success?” (NSF PAPPG 24-1, Chapter III). A design that names alternatives considered, states the evidence behind the chosen approach, and sets thresholds for when it would adapt reads as one that has been operated. A design with no contingency reads as one that has not.

How does a proposal prove the organization can execute?

A grant proposal proves the organization can execute through artifacts a third party could verify: program data, audited financials, prior award numbers, named staff with percent effort, and signed commitments. Demonstrating organizational capacity sets out which claims count as evidence and which read as assertion.

Federal agencies assess capacity twice. Peer reviewers score the narrative, and a separate administrative review — which the applicant never sees — weighs financial stability, quality of management systems, history of performance on prior awards, audit findings, and the ability to implement statutory requirements (2 CFR 200.206). The narrative and the back-office reality therefore have to match, because two different offices check them.

Newer organizations are not excluded, but they win by borrowing capacity rather than claiming it: a fiscal sponsor, a partner holding the functions they lack, adjacent track record attributed to named staff, or a first phase at a scale they have demonstrably run. The failure mode is the reverse — a grant proposal that jumps an order of magnitude beyond anything the applicant has operated, with no intermediate stage and no named mitigation.

What survives after the grant ends?

A sustainability plan names the durable asset, prices its recurring annual cost, and assigns that cost to a specific budget, revenue stream, or partner. Writing a sustainability plan treats the question as the risk assessment it is: funders are pricing the chance that their money produces an effect that disappears.

“We will seek additional funding” is the weakest available answer to a sustainability question, because it restates the funder’s question as the applicant’s plan. Nothing in the sentence is verifiable. Institutionalization into an existing budget, a secured policy or reimbursement code, earned revenue, partner absorption, and persistent capacity are all mechanisms a grant proposal can evidence with a letter, a rate, or a line item.

Some projects should end on schedule, and saying so is defensible. A demonstration that establishes an effect, a planning grant that produces an adopted plan, and a campaign that secures a rule change each complete their purpose inside the period of performance. The honest answer for those describes what persists — the evidence, the plan, the rule — rather than how the staff stay employed.

Why does the grant abstract carry disproportionate weight?

The grant abstract carries weight out of proportion to its length because it is read by everyone and used before anyone scores. The abstract and project summary covers what belongs in it, how funder specifications differ, and why it is drafted last.

Three uses happen before evaluation begins. Agencies route applications to panels and reviewers partly on abstract vocabulary, program staff triage large pools from summaries, and federal project summaries are published as public records once an award is made. Terminology borrowed from an adjacent field can send a grant proposal to reviewers who do not recognize its contribution.

The University of Michigan’s proposal guide states the consequence directly: “Some reviewers read only the abstract, e.g., a foundation board of directors’ member who votes on final funding decisions. Thus it is the most important single element in the proposal.” The same guide supplies the sequencing rule — “Though it appears first, the abstract should be edited last, as a concise summary of the proposal” (University of Michigan, Proposal Writer’s Guide). Writing it last also works as a completeness check: a summary that cannot state an objective without inventing one has found a hole in the narrative.

When does a letter of inquiry come before a proposal?

A letter of inquiry comes first when a private funder uses one to screen for fit before inviting a full application. The letter of inquiry covers the structure, the word budget per section, and the federal analogues that behave differently.

The letter compresses the same burdens into one to three pages, with one change of order: the amount, duration, and support type belong in the opening paragraph rather than the closing one. A program officer who cannot see the size of the request cannot triage it. Candid’s guidance names the difficulty plainly — “An effective LOI is often more difficult to write than a full proposal” (Candid, What to Include in Letters of Inquiry).

Federal preliminary documents share initials and almost nothing else. A letter of intent is advisory and helps an agency plan peer review; a preliminary proposal at the National Science Foundation can decide eligibility to submit a full grant proposal at all. Both are compliance requirements when a notice names them.

How do grant reviewers actually read a proposal?

Grant reviewers are subject-matter peers, usually unpaid volunteers, working through a stack against published criteria on a deadline — most spending hours rather than days on any single application. Writing for the reviewer assembles the survey and replication evidence on how that reading actually goes.

Two findings change how a grant proposal should be built. Assigned reviewers argue your case to a panel that has mostly skimmed, so the document has to hand them a portable one-paragraph claim. And reviewer agreement on strong applications is close to zero: a full replication of the National Institutes of Health process found that two randomly selected ratings of the same application were, on average, no more similar than two ratings of different applications (Pier et al., Proceedings of the National Academy of Sciences).

The University of Michigan’s guide sets the right reader model: “your reader is a busy, impatient, skeptical person who has no reason to give your proposal special consideration.” What follows is structural, not stylistic — mirror the funder’s headings, front-load every unit of text, spell out acronyms in each section, and allocate pages by published point weight. The National Institutes of Health gives the reason: “Reviewers are accustomed to finding information in specific sections,” which “saves reviewers from hunting” (NIH, General Grant Writing Tips).

How do you keep a grant proposal compliant?

A compliance matrix keeps a grant proposal compliant by converting every stated requirement into a numbered row with a source location, an owner, a place in the draft, and a status. Building a compliance matrix covers the extraction passes, the row fields, and who runs the final check.

Compliance is judged before competitiveness and by different people. The National Science Foundation publishes explicit grounds for returning a proposal without review, including one that “does not meet NSF proposal preparation requirements, such as page limitations, formatting instructions, and electronic submission,” one that “is not responsive to the NSF funding opportunity,” and one that “does not contain each of the required sections of the proposal” (NSF PAPPG 24-1, Chapter IV).

Purdue’s writing lab states the general consequence: “many grantmakers will not read past the point of your departure from the application rules, no matter how worthy the project is,” and “not following directions indicates carelessness—which is not a characteristic of a promising proposal” (Purdue OWL, Grant Writing). Compliance failures do not cause most declines. They cause most avoidable ones.

What should you do when a grant proposal is declined?

After a grant proposal is declined, request whatever feedback exists, weight each criticism by how many reviewers raised it, decide deliberately whether to resubmit, and harvest the reusable material. What to do when a grant is declined covers the resubmission rules and the evidence on what improves odds.

A decline reports how one panel scored one submission on one day. Federal agencies generally produce written critiques; most private funders do not, and a decline that says other projects offered a stronger fit contains no diagnostic information about the document at all.

Six assets survive intact and should be filed where the next application can find them: the need statement and its data, the project design, the budget and its assumptions, partner letters, the capacity section, and the reviewer comments themselves. The expensive failure is archiving a declined grant proposal under the competition that declined it and rebuilding the same material from scratch for the next deadline.

How the pieces fit together

The sections of a grant proposal are not modules; they are links in a single argument, and each one constrains the next. The table below maps the six burdens of proof to the section that carries each and the hub where it is covered in depth.

Burden of proof	Section that carries it	Where it is covered
Is the problem real?	Statement of need	This hub
Does the solution follow?	Objectives and project design	This hub
Can the applicant execute?	Organizational capacity	This hub
Will anyone know if it worked?	Evaluation plan	Evidence hub
Is the money right?	Budget and budget narrative	Budgets hub
Does the work survive?	Sustainability plan	This hub

The constraints run forward. The gap documented in the need sets the baseline every objective is measured against. The objectives determine which activities the design must contain, and the design’s dosage determines whether the objectives are arithmetically reachable. The staffing in the design determines the personnel lines, which is why writing a budget narrative is a derivation exercise rather than a pricing one. The objectives also determine the indicators, which is why writing an evaluation plan belongs beside the objectives rather than after them, and why a logic model drafted early usually produces cleaner objectives.

What reviewers actually catch is the break in that chain. A synthesis of decline reasons from inside the National Science Foundation lists misalignment as its own failure category, separate from weak framing and weak methods: “Sound theoretical framework, solid methodology, but they don’t align with each other” (Hapke, How NSF Makes Awards, UC Irvine Research Development). Two strong sections that do not connect score worse than two adequate sections that do.

Inconsistency is also the cheapest defect to find. A coordinator named in the design with no salary line, a participant count that differs between abstract and budget, a partner contribution that appears in a letter and not in the match — each is visible in a single pass. The Office of Justice Programs states the standard in one sentence: “Your budget should justify all expenses and be consistent with the program narrative” (OJP, Develop a Budget). Run the trace both ways before submission: forward from every objective to its activities, staff, cost, and measure, and backward from every budget line to the objective it serves.

Frequently asked questions

In what order should you draft a grant proposal?

Draft in dependency order, not document order: need, objectives, design and evaluation together, capacity, budget, sustainability, then the abstract. Building a compliance matrix precedes all of it. Drafting in page order produces objectives invented before the need is documented, which is the misalignment reviewers are trained to notice.

How long does it take to write a grant proposal?

A first federal application typically takes weeks of elapsed time across several people, most of it spent on evidence gathering, partner commitments, and budget derivation rather than prose. Repeat competitions are faster because the compliance rows, capacity section, and data citations already exist. Registration status is the constraint least amenable to a rush.

Can you reuse a grant proposal for a different funder?

Reuse the research, not the packaging. The need data, project design, budget assumptions, and capacity evidence transfer across funders working the same problem. The fit statement, section headings, requested amount, and every response written to a specific criterion do not. A recycled proposal reads as generic and is declined on responsiveness.

What happens when a funding announcement contradicts itself?

Treat the contradiction as an open row with a named owner and ask the agency contact in writing before the question deadline. Agencies must make clarifying information available to all potential applicants, so answers are typically posted publicly. A citable answer is worth far more than an internal guess about which instruction governs.

Who should run the final review of a grant proposal?

Someone who did not write the draft, working from the compliance register rather than the document. Authors verify what they intended to write; a fresh reader verifies what is on the page. Schedule the check far enough before the deadline that a failed requirement can still be remediated rather than merely discovered.

Do foundation proposals use the same sections as federal ones?

Yes, compressed. A foundation asking for a short narrative still expects need, plan, capacity, measurement, and budget, and often screens through a letter of inquiry before inviting anything longer. What changes is enforcement: federal criteria are published and scored, while a foundation program officer’s judgment is not constrained by a rubric.

Related topics

• Grant Funding Fundamentals — how the funding system works
• Finding and Qualifying Funding — search, funder research, and go/no-go
• Eligibility and Organizational Readiness — registrations, documents, and internal controls
• Budgets and Grant Finance — cost categories, indirect rates, and cash mechanics
• Evidence, Evaluation, and Data — logic models, indicators, and evaluation design
• Funding Tracks by Funder Type — federal, foundation, corporate, and research funding
• Managing the Award — reporting, monitoring, audit, and closeout

Sources

1. U.S. Office of Management and Budget. Appendix I to Part 200 — Full Text of the Notice of Funding Opportunity. eCFR. https://www.ecfr.gov/current/title-2/subtitle-A/chapter-II/part-200/appendix-Appendix%20I%20to%20Part%20200 (accessed 2026-08-11)
2. U.S. Department of Education. 34 CFR 75.210 — General selection criteria. Cornell Legal Information Institute. https://www.law.cornell.edu/cfr/text/34/75.210 (accessed 2026-08-11)
3. U.S. Office of Management and Budget. 2 CFR 200.206 — Federal agency review of risk posed by applicants. eCFR. https://www.ecfr.gov/current/title-2/section-200.206 (accessed 2026-08-11)
4. U.S. National Science Foundation. PAPPG 24-1, Chapter III — Proposal Processing and Review. https://www.nsf.gov/policies/pappg/24-1/ch-3-proposal-processing-review (accessed 2026-08-11)
5. U.S. National Science Foundation. PAPPG 24-1, Chapter IV — Non-Award Decisions and Transactions. https://www.nsf.gov/policies/pappg/24-1/ch-4-non-award-decisions-transactions (accessed 2026-08-11)
6. Office of Justice Programs. Develop a Budget. U.S. Department of Justice. https://www.ojp.gov/funding/grants101/develop-budget (accessed 2026-08-11)
7. University of Michigan Office of Research and Sponsored Projects. Proposal Writer’s Guide. https://sites.lsa.umich.edu/nextprof-science/wp-content/uploads/sites/130/2018/04/3-University-of-Michigan-Proposal-Writers-Guide.pdf (accessed 2026-08-11)
8. Pier, E.L., Brauer, M., Filut, A., et al. Low agreement among reviewers evaluating the same NIH grant applications. Proceedings of the National Academy of Sciences 115(12), 2018. https://www.pnas.org/doi/10.1073/pnas.1714379115 (accessed 2026-08-11)
9. Purdue Online Writing Lab. Grant Writing. Purdue University. https://owl.purdue.edu/owl/subject_specific_writing/professional_technical_writing/grant_writing/index.html (accessed 2026-08-11)
10. Hapke, Holly M. How NSF Makes Awards: Insights from Inside the Black Box. University of California, Irvine, School of Social Sciences Research Development, 2021. https://www.researchdevelopment.socsci.uci.edu/files/documents/2021/How%20NSF%20Makes%20Awards.pdf (accessed 2026-08-11)
11. Candid. What to Include in Letters of Inquiry. Candid Learning. https://learning.candid.org/what-to-include-in-letters-of-inquiry (accessed 2026-08-11)
12. National Institutes of Health. General Grant Writing Tips. https://grants.nih.gov/grants-process/write-application/general-grant-writing-tips (accessed 2026-08-11)