The most common way small businesses build a budget is to take last year's numbers, add a percentage to revenue, add a smaller percentage to expenses, and call it a plan.
It is fast and it feels reasonable. It also carries every one of last year's assumptions forward without anyone looking at them, including the ones that stopped being true.
I prefer to build the budget from the bottom. It takes longer. It is also the only version of the exercise that tells you anything you did not already know.
Start with drivers, not totals
Revenue is not a number. It is the result of a few things multiplied together. In staffing, that might be the number of active clients, the hours they order, and the margin on each hour. In a software studio, it might be the number of engagements, their average length, and how much of the team's time is billable.
Every business has a small set of drivers like that. Write them down, estimate each one for next year, and let the revenue number fall out of the math. When the number looks wrong, you will be able to see which assumption is carrying it.
That is the real benefit. A top-down number gives you a target. A driver-based number gives you a model you can argue with.
Every expense line has to earn its place again
On the cost side, I want each recurring expense justified as if we were signing up for it today. Software subscriptions. Vendor contracts. Roles that were created for a reason that may no longer apply.
Most of the time the line survives. Sometimes it does not, and that is where the savings live. Not in heroic cost-cutting, but in the quiet accumulation of things nobody has questioned since they were set up.
The same goes for headcount. Rather than budgeting for last year's team plus raises, I want to know what work next year actually requires and who is going to do it. That conversation usually surfaces one role that is overloaded and another that has drifted, and both are better addressed in a planning meeting than in a crisis.
Budget the plan, not the hope
A budget should fund the decisions you made in planning. If the plan says we are launching something in the second quarter, the budget needs the people, the marketing, and the time to get there, along with a realistic view of when revenue from it will start.
The common mistake is to budget the cost of a new initiative and also assume it starts paying back in the same quarter. It rarely does. New things take longer than planned and earn less at first than anyone hoped. Budgeting as if they will not is how a reasonable plan turns into a cash problem by summer.
Build three versions
I like to see three cases: the one we expect, a weaker one, and a stronger one. Not as a forecasting exercise, but as a decision exercise.
For the weaker case, the question is what we would cut, in what order, and at what trigger. Decide that now, while you are calm. If revenue drops below a certain line for two consecutive months, these specific things pause. Writing that down in October means you do not have to make the decision under pressure in April.
For the stronger case, the question is where the extra cash goes. If you have not decided, it tends to get absorbed by whatever is loudest at the time.
Separate operating cash from reserves
The budget should show what the business needs to operate and, separately, what it should hold back. A reserve that is mixed into the operating account tends to get spent, not because anyone is careless, but because it looks available.
How large that reserve should be depends on the business. A company with lumpy receivables needs more than one with predictable monthly billing. The point is to decide the number on purpose and keep it somewhere it is not tempting.
Review it monthly, not annually
A budget that only gets opened in December is not doing its job. I want a short monthly comparison of actuals to plan, looking at the drivers rather than the totals. When something moves, we want to know which assumption broke and whether it changes anything we decided.
That review is where the bottom-up approach pays for itself. When you know what built the number, you know where to look when the number is off.
It is more work than adding ten percent. It is also the difference between a budget you own and one that owns you.