Most businesses that tell us they have a budget mean they built one, once, at the start of a year. It was probably a reasonable piece of work. It also stopped affecting any decision roughly six weeks later, because nothing was ever compared against it.
The three ways budgets die
It was last year plus a percentage
The fastest way to produce a budget is to take last year's actuals and add a growth assumption. It's also the fastest way to bake in every inefficiency you already had, including the subscriptions nobody uses and the service line that loses money. A budget built this way can't reveal a problem, because the problem is in the baseline.
Seasonality got averaged away
Dividing an annual figure by twelve produces a budget that's wrong every single month in a business with any real seasonality — too generous in the quiet months, too tight in the busy ones. After two or three months of meaningless variances, people quite reasonably stop looking at the report.
Nobody ever compared it to anything
This is the big one. A budget's entire value is as a benchmark. Without a monthly comparison against actuals, it's a prediction nobody checked — and an unchecked prediction can't tell you that payroll is drifting or that a category is running hot.
What a useful monthly review looks like
- Close the month properly — variance reporting off unreconciled books is noise.
- Compare every material line to plan, not just the bottom line.
- Trace each significant gap to a cause, rather than logging the number.
- Separate real drift from timing differences, so the report stays short.
- Where reality has genuinely moved, change the plan instead of defending it.
When to revise versus when to hold
Owners often ask whether they should update the budget mid-year. The test we use: has an assumption changed, or has performance changed? A new lease, a lost major client, or a price increase from a key supplier are changed assumptions — the budget should reflect them. Simply missing the target for three months is changed performance, and rewriting the budget to match it just removes the thing that was telling you something.
A budget you keep revising downward until it matches actuals isn't a budget. It's a transcript.