Most startup budgets look great in January and break by March. The reason is almost always the same: the plan started with a revenue number someone hoped to hit and worked backward. A budget that holds up starts with what you actually control and builds up from there.
1. Start with cash and runway
Before anything else, know how much cash you have and how fast you're spending it. Your runway is your cash balance divided by your monthly net burn, which is cash going out minus cash coming in. Every decision in the budget should be checked against what it does to that number.
2. Build revenue from drivers, not percentages
"Grow 20% a month" isn't a plan. Build revenue from the things that produce it: how many leads you expect, how many become customers, what they pay, and how long they stay. When results come in, you'll know exactly which assumption was off.
3. Let headcount drive expenses
For most startups, people are the largest cost. List every planned hire with a start month and salary, then add the real cost on top: payroll taxes, benefits, equipment, and software. A hire in month three costs very different money than the same hire in month nine.
4. Plan everything else on purpose
Group the rest into a few categories, such as software, marketing, rent, and professional fees. Tie marketing spending to the customer targets it's meant to deliver, so you can tell whether it's working.
5. Plan for three outcomes
Build a base case, a downside case, and an upside case. Decide ahead of time what you'll do if the downside happens: which hires get delayed, which spending gets cut, and when you'd need to raise money. Making those calls before you're under pressure leads to better decisions.
6. Compare and update every month
A budget only helps if you use it. Compare actual results to the budget every month, explain the big differences, and update your forecast each quarter. The budget is the plan you set at the start of the year. The forecast is your best view of where you're actually headed.
Common mistakes to avoid
- Revenue that jumps sharply with nothing in the plan to explain it
- Forgetting payroll taxes and benefits on top of salaries
- Assuming customers pay the day you invoice them
- Leaving out one-time costs like legal fees, recruiting, or equipment
- No cushion for the things you can't predict
We'd build the downside case first. If the business survives that version, everything better is upside rather than a bet.