The budget is read by an accountant before it is read by a strategist.

Two readers, years apart

The first is the evaluator, scoring whether the budget is plausible and cost-effective against the activities proposed. The second arrives after the award: a controller checking whether each cost was eligible, evidenced and inside the period.

Applications are usually written for the first reader. Problems are almost always created for the second.

The lag nobody budgets for

Public funding is usually reimbursed after the cost, against a claim, sometimes months later. A project that is fully funded on paper can still run out of cash in month five, and the cash requirement is a fact about your balance sheet rather than about the grant.

Five layers

• Activity-based budget. Every cost line traceable to a work package, so the panel can see what the money does.
• Programme cost categories. Personnel, subcontracting, equipment, travel, overheads, each in the structure and unit basis the call defines.
• Co-funding plan. The source, the timing, and the evidence that will prove it, since a promise of co-funding is not co-funding.
• Cash flow through the period. Spend curve, claim dates, reimbursement dates, and the headroom required at the worst month.
• Audit trail map. For each category, what document proves the cost, who signs it, and where it is kept.

Cost rules decide more than arithmetic

Unit rates versus actual costs, flat-rate overheads, eligibility windows, procurement thresholds, rules on costs incurred before signature. These determine the shape of the budget before a single number is entered, and a model that ignores them produces a figure that cannot be claimed.

The familiar mistakes

• Building the budget to hit the grant ceiling, then inventing activities to justify it.
• Overheads allocated by a method nobody can explain to a controller.
• Co-funding promised from revenue that has not yet been earned and cannot be evidenced at claim.
• No headroom for the reimbursement lag, so the project is funded and illiquid at once.
• Unit rates taken from memory instead of from the published tables.
• Costs placed in a category for convenience, then defended during audit.
• Partners counting the same cost, or the same staff month, in two budgets.

Each is a scoring risk at application and an audit finding later. The second is worse than the first.

What the panel actually tests

Whether the cost per unit of output is defensible against comparable projects, and whether the team understands the funding mechanics. A budget that shows the lag is understood reads as experienced.

The budget set

• Budget model in the programme's cost structure, linked to work packages and effort.
• Cost justification narrative: what each category buys, on what basis, at what rate.
• Co-funding plan with source, timing and the evidence attached to each contribution.
• Cash-flow and headroom view across the project period, showing the worst month.
• Partner allocation sheet, so no cost is counted twice and each share is traceable.
• Audit trail map, written for the controller who arrives after everyone has moved on.

What we need from you

The call's cost rules in full, current payroll and rate data, supplier quotes where procurement applies, and the co-funding positions you can actually evidence.

Rules, then model, then test

Read the cost rules and the eligible categories before modelling anything.

Build the budget from the work programme rather than the reverse, so every line has an activity behind it.

Then a rehearsal audit: we ask the questions a controller would ask, line by line, and change what does not survive.

Who does what

Your finance team owns the accounting records, statutory treatment and the actual claims. We own the model, its structure and its defensibility. Tax and grant-compliance opinions sit with your advisers.

Where the work stops

It is not an audit opinion and does not guarantee eligibility of a cost, which is decided by the funder.

It does not produce the accounting evidence, only the map of what that evidence must be.

It does not replace the application itself: grant business plan.

And it does not run the post-award reporting cycle, which is a recurring obligation rather than a document.

Related work

Where the same numbers have to convince an investor instead of a funder: startup financial model.

Where this work leads

One document answers one reader, but the question behind it rarely stays inside a single direction.

• Project & Investment Strategy

The level above

Documents solve a defined file. Where the decision itself is open, the work moves up a level, into a direction, and eventually into growth strategy.