Four questions every budget line must answer
Reviewers do not read budgets as accounting documents. They read them as a test of whether the program plan is real. A line passes that test when it answers four questions on its face.
What is the unit and quantity? Miles, hours, FTE share, participants, sessions. A round number with no unit (“Supplies: $5,000”) announces that nobody did the math. The same amount as “40 student kits x $95 + 2 printer refills x $620” is a plan.
Where does the rate come from? A salary from your payroll, a mileage rate from the current IRS standard, a vendor quote, a published license price. The rate source is what turns an estimate into evidence, and it is the first thing a finance reviewer asks for.
Which activity does it serve? Every line should trace to something the narrative says the program will do. If you built a logic model, its inputs column is the skeleton of this budget, and a cost with no activity behind it is a flag in both directions: either the budget is padded or the narrative is incomplete. The narrative side of that trace, section by section, is laid out in the grant proposal template.
What does the funder’s rule say? Salary caps, indirect cost limits, disallowed categories, match requirements. The live opportunity governs, and Candid’s proposal budgeting guidance is a sound general reference for how foundation reviewers expect budgets and narratives to fit together.
The template carries all four as columns, so a blank cell is a visible unfinished task rather than a silent gap.
Grant budget templateLine-item budget with columns for unit, quantity, rate, calculation to run, activity trace, funder rule check, and narrative section, plus worked example rows for personnel and fringe. Formulas are written out so they survive any spreadsheet tool.
CSV templateStaff time math without double counting
Personnel is usually the largest category and the one most often built backwards, starting from a dollar amount that feels fundable instead of from time the program actually needs.
Worked example
Pricing a program coordinator at 0.4 FTE, with the double counting check
Assumptions, all declared: a fictional organization pays its program coordinator a $52,000 annual salary; its written fringe benefit rate is 24 percent of salary; the grant period is 12 months; a full-time year is 2,080 hours.
Step 1: estimate the time from the activity plan, not from the budget target. Two tutoring sessions a week with preparation, volunteer coordination, family nights, and data entry come to roughly 16 hours a week. 16 / 40 = 0.4 FTE.
Step 2: salary charged to the grant = $52,000 x 0.40 = $20,800.
Step 3: fringe = $20,800 x 0.24 = $4,992. Fringe is calculated on the salary charged to this grant, never on the full salary.
Step 4: the double counting check. The coordinator is already charged at 70 percent effort to another funded project. 70 + 40 = 110 percent of one human. That budget is unbuildable, and cross-checking effort across all active and pending grants is the single most valuable personnel review a small nonprofit can run. The fix is honest: reduce this grant’s share to 0.30 FTE ($52,000 x 0.30 = $15,600, fringe $3,744) and either shrink the activity plan to match or add a second funded role.
Sanity check in hours: 0.4 FTE = 832 hours across the year, which the 16-hour weekly estimate supports. Total personnel for the role at the corrected 0.3 FTE: $15,600 + $3,744 = $19,344.
The pattern generalizes: salary x FTE share, fringe rate x charged salary, and one organization-wide effort ledger that keeps every person’s total at or under 100 percent. When a person is split across funders, the split must also match how their time will actually be tracked, because the grant report will have to reconcile with timesheets.
Indirect costs and the de minimis rate
Indirect costs pay for the administration that makes programs possible: bookkeeping, audit, rent for the office that is not program space, the director’s time on compliance. Two failures are common: claiming nothing, which quietly starves the organization, and applying federal rules to funders they do not bind.
The working rule for a small nonprofit: on a federal application with no negotiated rate, elect the de minimis rate, compute the MTDC base explicitly in the budget file, and show the multiplication. On a foundation application, read the funder’s policy first and never write “15 percent per 2 CFR 200.414” to a funder the regulation does not bind; cite their own policy or ask. In every case the indirect line follows the same discipline as any other line: a rate, a base, a source for both.
Reconciliation: one story across four documents
A budget that is internally correct can still sink a proposal by disagreeing with the documents around it. Before submission, reconcile four ways:
- Budget vs. narrative: every activity in the narrative has money behind it, and every budget line has an activity in front of it. Twelve tutors in the story and ten stipends in the spreadsheet is the kind of contradiction reviewers remember.
- Budget vs. staffing: names and effort levels in the narrative match the personnel lines exactly.
- Budget vs. timeline: costs land in the periods where their activities happen. In a multi-year budget, resist dividing the total evenly across years; year one carries startup and training, later years carry full delivery, and a flat split tells the reviewer the schedule was never costed. Declare any inflation assumption on multi-year salaries instead of hiding it in the rates.
- Budget vs. budget narrative: the narrative explains basis and purpose (“0.3 FTE of the coordinator’s $52,000 salary, to run 56 tutoring sessions”), it does not restate totals. If the number and its explanation live in different documents, keep one source file and generate both from it. The complete category-by-category worksheet for that document is the budget narrative template.
The same reconciliation sweep belongs in the coherence gate of the grant submission checklist, where it catches the drift that accumulates during last-week edits. In-kind contributions, where the funder permits them, follow the identical rule: a documented value basis and a named commitment, never an inflated market guess.
Our position on padded budgets
A clean budget cannot guarantee funding, and no template can promise a funder’s decision. What it does is remove the most checkable reasons for doubt, and it leaves you with a file you can actually manage the grant against after an award. The evaluation section deserves the same treatment as every other cost: hours and rates in a small work breakdown, as covered in the evaluation plan template, not a decorative percentage.