If you closed a round this year, the 2027 budget matters more than you think. Before the raise, the budget was you and a spreadsheet. Now it's an investor commitment and a management contract, and they don't want the same number. The first budget after a raise is where most founders get the shape of it wrong.
Most founders think about hiring, not budgeting. But a manager without a budget still comes to you for every spend decision, every trade-off, every conversation about headcount, every "can we afford this". You stay the bottleneck. The whole point of a budget is to hand your people money and constraints so they can decide without you. These are the things I see founders get wrong.
1. Decide what the budget is for before you build it
Is it a target for the team, a commitment to investors, a cash plan, or a stretch goal? Each one wants a different number. Most founders build one and try to use it for all four, which is why the same document gets defended in January and ignored by March. Pick the primary purpose: everything else follows from that.
2. Build it bottom-up from the operating plan, not top-down from a growth rate
"40% growth" is not a plan, it's a headline. Start from what the team will actually do next year: which products, which markets, which channels, which customers. Then check that the activity ladders up to the revenue number. If it doesn't, the number is fiction.
3. Your managers do the first draft, you set the constraints
This is the bit founders skip and then wonder why they did the whole budget themselves at 1am. Managers need to build their own numbers; sales builds sales, product builds product. But they need your constraints going in; otherwise they'll build a wishlist with no connection to reality. While they own the numbers, you own the shape of the whole company, and their department has to fit into it. At the same time, managers who build their own budget defend their own budget. Managers who receive a budget resent it.
4. Budget headcount by workload, not by wishlist
"Hire a Head of Sales in Q2" is a wish. Work out what actually needs doing, when, and roughly how many hours it takes. Then decide who does it. Wishlist hires either die before Q2 or come back stronger with real justification behind them.
5. Do it in cash, not just P&L
Revenue recognition isn't cash and neither are your cost accruals; VAT, working capital, and corporation tax all eat or free up cash in months when the P&L looks fine. Build a cash version alongside the P&L; it's the one that tells you when you run out. If you can't remember which one your last budget was, it was the P&L.
6. Build three versions, and actually use the downside
Base, upside, downside. The downside isn't there to be pessimistic; it's there to give you a trigger: if revenue is below X by end of Q2, we do Y. Founders gravitate toward the upside and ignore the downside until they're in it. Reverse that habit, because the downside is the version that saves you.
7. Share the draft with your board in November, not on Christmas Eve
The worst outcome is a budget you defend for two hours in a January board meeting and rebuild in February. Give the board the draft in November; ask them what they'd challenge, then bake that in before the version they sign off on. You want the board to feel they helped shape it, not that they inherited it.
8. If you're not comparing to actuals monthly, you're forecasting not budgeting
The budget only earns its keep when you sit it next to the actual. Every month. If you know you won't do that, save yourself November and build a rolling forecast instead; both are valid, but a budget without monthly review isn't a budget.
The short version
Decide what it's for. Build it bottom-up, with your managers drafting their own numbers under real constraints. Run a cash version alongside the P&L. Build three scenarios and pre-agree what triggers the downside. Share with the board in November. Compare to actuals monthly, or call it something else.
The founders who do this well stop being the finance bottleneck by Q2. The ones who don't spend 2027 approving £2k spend decisions on a Thursday night.
