Bookkeeping · Financial Statements · Brief · Working level
Budget vs. actual reports: building the comparison and chasing the right variances
How to set up a usable small-business budget-versus-actual report, and the threshold rules that separate variances worth investigating from noise.
A budget-versus-actual report is the ordinary P&L with three columns bolted on: what you planned each line to be, what it was, and the difference. Its value is speed — instead of reading every number, you read only the surprises. Most months, a well-built report lets you clear the review in ten minutes because most lines land where planned.
Building the comparison
- Budget at the level you manage, not the level you record. Ten to twenty summary lines — revenue by stream, COGS, payroll, rent, marketing, and so on — not eighty subaccounts. Detail belongs in the ledger; the budget is for decisions.
- Spread annual figures by realistic month. Seasonal revenue budgeted as one-twelfth per month generates phantom variances all year. Use last year's monthly shape as the spreading pattern.
- Enter the budget into your ledger software so the report generates itself; a budget living in a separate spreadsheet dies by March.
- Run two versions monthly: the month alone, and year-to-date. YTD is the corrective lens — a bad month inside a on-plan year is a different situation from a bad month extending a bad trend.
If you have no budget, run the identical report against last year's actuals. The common-size statement is the percentage cousin of this comparison; use both.
The threshold rules
The failure mode of variance review is treating every difference as a question. Use a two-gate filter:
Investigate a line only if it fails both gates (illustrative thresholds for a business around $50,000/month of revenue):
| Gate | Threshold | What it screens out |
|---|---|---|
| Percentage | Variance > 10% of budget | Normal wobble on large lines (payroll ±2%) |
| Dollar floor | Variance > $500 | Large percentages on trivial lines (bank fees +80% = $40) |
| Direction rule | Same-direction miss 3 months running | Slow drift that never trips a single-month gate |
Scale the dollar floor with the business — roughly 1% of monthly revenue is a defensible default. The third rule matters most: a marketing line running 6% over plan every single month never trips the 10% gate but is a genuine trend, the same slow-drift problem common-size reading targets.
Investigating a variance that matters
For each flagged line, answer three questions in order:
- Is it real? Check for misposted or duplicated transactions first — a meaningful share of "variances" are bookkeeping errors, which makes this report a free audit of the ledger.
- Is it timing? An annual insurance bill budgeted in May but paid in June creates two offsetting fake variances. Note it and move on.
- Is it a decision? Only what survives the first two questions is management information: a supplier raised prices, a hire came early, revenue is genuinely soft. Each of these ends in an action or a conscious budget revision.
Two boundary notes. Budgets are management tools, not tax documents — the IRS cares about your actual books and records (the standard set out in Publication 583), and no budget figure belongs anywhere near a return. And if a variance investigation turns up spending that changes your estimated-tax picture mid-year, the self-employed tax center covers adjusting quarterly payments. The report itself belongs in the standing monthly stack described in the monthly reporting package.
Frequently asked questions
- What is a budget vs. actual report?
- A P&L-format report with three extra columns: the budgeted amount for each line, the actual result, and the variance between them (in dollars and percent). It converts the monthly review from 'what happened?' to 'what happened that we didn't expect?' — a far faster question to answer.
- Which budget variances should a small business investigate?
- Use a two-part threshold: investigate a line only when the variance exceeds both a percentage (commonly 10%) and a dollar floor (commonly $250–$500, scaled to the business). The dollar floor screens out large percentages on tiny lines; the percentage screens out small wobbles on large lines.
- Do I need a formal budget to do variance analysis?
- No. If you have no budget, use last year's same-month actuals as the baseline — the comparison mechanics and thresholds work identically. A prior-year comparison catches most of what a budget catches; the budget adds intent, letting you test the plan rather than just the past.