A budget is a plan for what the charity intends to do next year and what that will cost. It is not a report, a forecast, or a tidied-up version of last year. That distinction matters because it decides how you build it, and building it the wrong way produces a document the board cannot use for anything except comparison.
This is for a treasurer or chief executive in front of a blank spreadsheet, or a board that has inherited a format from someone who left and nobody has questioned since.
If the budget keeps failing because the plan behind it was never settled, strategy support is the conversation before this one. What follows assumes you know roughly what the charity intends to do.
Start from activities, not from last year
The common method is last year’s figures plus a percentage. It takes twenty minutes, it is easy to defend in a meeting, and it carries forward every mistake the charity has ever made – the post that was underfunded from the start, the programme nobody has evaluated since 2022, the venue hire that stopped being good value two years ago.
The alternative is to start from what the charity intends to do. List the activities for the year. For each one, work out what it requires: staff time, materials, venue, travel, printing, the share of the coordinator’s week it takes up. Add them up. That total is your expenditure budget, and every line in it traces back to something the charity decided to do.
The first year of doing it this way is slow. Every year after is a revision rather than a rebuild.
What makes it worth the effort is a question the board will eventually ask. If income fell twenty per cent, what would we stop. A budget built from activities answers that in an afternoon, because the costs are attached to the things they pay for. A budget built from last year plus five per cent cannot answer it at all, and the board ends up cutting whatever is easiest to cut rather than whatever matters least.
The costs nobody counts
Direct costs are the straightforward half. The sessional worker, the room hire, the materials for the group.
The half that gets missed is everything holding the charity up. The coordinator’s hours spent on admin, reporting and email. Insurance. The accountant or independent examiner. Software subscriptions that renew quietly. Bank charges. Training. The trustee who has never claimed a mileage expense in four years and whose successor will.
A budget that leaves those out describes a charity that does not exist, and it produces a surplus that disappears the moment someone counts properly. Each activity should carry a fair share of them. How you apportion that share is a judgement rather than a formula – by staff time, by headcount, by proportion of direct costs – and the important part is choosing a basis and applying it consistently rather than finding the perfect one. On the funder-facing side of the same problem, why winning grants can leave your charity worse off covers what happens when overheads are left out of a bid.
Income, and the difference between a plan and a hope
Split income into three groups and keep them apart on the page.
Confirmed. The money is contracted, granted, or in the bank. Second, likely – a grant you have applied for with a reasonable history, a contract expected to renew, an event that has run for six years. Put a probability against each and be honest about it. Third, hoped for. The application not yet written. The corporate partner; someone knows someone at.
The test of a budget is what sits in that third group. If the gap between planned costs and confirmed income is being closed by hoped-for money, the budget is not a plan. It is a wish with a spreadsheet around it, and the board is being asked to approve a level of spending the charity cannot fund.
Restricted and unrestricted income need separating too. A budget that nets them together will show a surplus the charity is not allowed to spend, which is a different problem from having no surplus and a more dangerous one, because it looks like good news. The distinction has to exist in the records before it can appear in the budget, which is where financial controls and planning meet.
One more thing to check before the budget is finished. Any funding that ends inside the budget year. A three-year grant finishing in month eight is not a problem for next year’s budget, it is a problem for this one, and the four months after it stops are the part the budget has to show.
Building a budget from activities for the first time is the hard year. After that it is a revision rather than a rebuild. If you want someone alongside you for the first pass, book a call and we can work through the activity list, the overheads, and what the board should be asked to approve.
When the cash actually arrives
A charity can be solvent across the year and unable to pay March payroll. The annual budget will not tell you that. It is the single most useful thing a small charity can add to its planning and one of the least common.
The exercise is to lay out, month by month, when money comes in and when it goes out. Grant instalments paid quarterly in arrears. An event that spends in February and raises in June. Invoices settled at sixty days by an organisation that treats that as standard. Payroll, which does not wait. Insurance renewing in one lump.
What comes out of it is a shape, and the shape has a lowest point. Knowing which month is tightest, before it arrives, is the whole benefit. A charity that knows February is the problem can move an invoice date, ask a funder about instalment timing, or delay a purchase. A charity that discovers it in February can do none of those things.
This also connects the budget to what the board reviews in-year. The quarterly questions in what trustees should ask about the money each quarter are far easier to answer when someone mapped the cash at the start.
What the board is approving
Approval gets treated as a formality more often than it should. The budget appears, the treasurer talks for five minutes, somebody proposes, somebody seconds.
Four questions change that. How were the income figures arrived at, and which of them are confirmed. What happens if the largest single income line does not come in. Are all the overheads in here, including staff time on admin. And what would we stop first if we had to stop something.
A treasurer who can answer those has built a budget the board can rely on. A treasurer who cannot has built a spreadsheet. Recording that the board asked and what it was told is the same discipline that applies to any trustee decision that might be examined later.
And the timing. The budget is approved before the financial year starts, not three months into it. A charity operating for a quarter on an unapproved budget has been spending against a plan nobody agreed.
Building in room to move
Contingency is worth having and rarely done honestly. A percentage line labelled contingency that has been spent by month four was never contingency; it was an underestimate with a friendly name.
The more useful version is deciding in advance what a smaller year looks like. Take the budget at ninety per cent of expected income and name what specifically comes out – not a percentage cut across every line, which is how charities end up doing everything badly. That decision is far easier to make in October with nothing at stake than in June with a funding gap. It is also where how much risk your board can accept stops being abstract, because the budget is where capacity turns into numbers.
Reserves belong in the same conversation. If the plan assumes drawing on them, the budget should say so explicitly, and the reserves policy should permit it. A planned draw is a decision. An unplanned one is a discovery.
When it stops matching reality
Every budget is wrong by month three. The question is what happens next.
Reporting variances against a budget that stopped being achievable in April generates noise for the rest of the year. Everyone knows the numbers are meaningless; the finance item becomes a ritual, and a real problem arriving in month nine looks like all the other variances.
There is a difference between a variance you explain and a plan you replace. A grant arriving late is a variance. A grant not arriving at all means the budget is describing a year that is not happening, and the board should be asked to approve a revised one rather than continue reporting against fiction.
Once or twice a year is usually enough. More than that, and the budget was never robust to begin with, which is a signal about how it was built.
The question to put to your board
Not whether the budget balances. Whether anyone in the room could explain how it was built, and what would come out of it first.
