The anatomy of a 100-point federal grant proposal

Start with the review criteria, not the narrative outline

Most federal discretionary programs publish their scoring model in the review section of the notice of funding opportunity. The criteria and their point values are the closest thing an applicant gets to the answer key. Yet many teams still open a blank document, write an introduction about the organization's history, and only check the criteria once the draft is finished. By then the page budget is spent on the wrong things.

A stronger habit is to treat the criteria table as the skeleton of the proposal. Before anyone writes a sentence, copy every criterion and sub-criterion into a working outline, note its point value, and record the exact NOFO section it points back to. HRSA, for example, has long organized many NOFOs around six criteria: Need, Response, Evaluative Measures, Impact, Resources and Capabilities, and Support Requested. Other agencies use different labels, but the idea is the same. Reviewers score against that list, in that order, and they look for your answer where the criterion tells them to look.

Allocate pages in proportion to points

If a criterion is worth 35 of 100 points and your page limit is 40 pages, it deserves roughly 14 pages of focused content. That arithmetic sounds obvious, yet proposals routinely give eight pages to organizational history worth 5 points and three pages to an approach section worth 30. Reviewers notice the imbalance even if they never say it out loud, because thin sections produce thin strengths in their written comments.

Build a page budget at the outline stage and hold writers to it. Adjust for content that naturally runs long, such as a work plan table, but write the adjustment down so the trade-off is deliberate. Check whether the page limit counts attachments as well as the narrative. In many HRSA NOFOs it does, and teams that discover this a week before the deadline end up cutting letters of support or staff biographies in a panic.

• List every criterion and sub-criterion with its point value
• Set a target page count for each, proportional to points
• Mark which attachments count against the limit
• Revisit the budget after the first full draft, not after the third

Mirror the criterion language in your headings

Objective reviewers work through a scoring form with a box for each criterion. They read your application with that form open beside it. When your heading says 'Section 3: Response, Work Plan' and the first paragraph restates what the criterion asks for, the reviewer can confirm the element exists and move on to judging quality. When the same content is buried under a creative heading like 'Our Path Forward', they have to hunt, and a tired reviewer at hour six of a panel week may simply note it as missing.

Mirroring does not mean copying the NOFO back to the agency. It means using the agency's nouns. If the NOFO asks for 'a description of how the applicant will engage community health workers in outreach to agricultural workers', your sub-heading should name community health workers, outreach and agricultural workers. The sentence underneath should then add something the reviewer could not have guessed: who, how many, by when, and with what evidence that the approach works.

Build an evidence plan for every scored claim

Every criterion invites claims, and every claim needs support. A need statement needs current data with a source and a year. An approach section needs a rationale, ideally drawn from published research or your own prior outcomes. A capacity section needs proof that you have managed federal money before: prior award numbers, audit results, staff who have run similar programs. The applicants who score well tend to treat evidence as a separate workstream, gathered before drafting starts.

A simple evidence plan is a table with three columns: the claim you need to make, the source that supports it, and where that source lives. If the source does not exist, you either go get it or you soften the claim. This is also where most fabrication risk sits. Under deadline pressure, a writer can be tempted to round a number or cite a statistic from memory. Reviewers with field experience catch these, and an inflated claim in the need section can cast doubt on everything that follows.

Do not leave easy points on the table

Some criteria are close to binary. Did the applicant include a logic model? Is there a staffing plan with position descriptions? Does the budget justification match the SF-424A line by line? These items are tedious, but they are among the cheapest points in the application. A proposal that misses two or three of them can lose more ground than a weak paragraph in the approach section.

Many NOFOs also list funding priorities or preference points, for example for applicants serving a defined high-need area or meeting a statutory preference. These are often awarded through a separate review step, and they require documentation the narrative alone will not provide. Read the eligibility and priority sections as carefully as the criteria, and assign one person to verify each requirement against the final package before submission.

Run an internal mock review before you submit

The single most useful quality step we have seen is a mock review by someone who did not write the proposal. Give them the NOFO scoring section, a blank scoring sheet and 90 minutes. Ask them to score each criterion and write two strengths and two weaknesses, exactly as a federal reviewer would. The gaps they find in 90 minutes are the gaps a real panel will find.

ELIXIR CONSUMERS builds this into the drafting workflow: each section is mapped to the rubric line it answers, and every factual sentence links back to the vault document it came from, so the mock reviewer can check sources without chasing the writer. Whatever tools you use, the principle holds. Write to the scoring form, budget pages by points, and prove every claim. That is most of what separates a fundable 88 from an unfunded 74.