Most businesses that build a budget also, eventually, run some version of a budget vs. actual report. Far fewer do anything with it. The report gets generated, glanced at, filed away — and the business keeps operating exactly as it was before. That's not a variance analysis problem, it's a process problem, and it's fixable.
What a Budget vs. Actual (Variance) Report Is
A variance report compares what you planned to spend or earn (the budget) against what actually happened (the actuals), for the same period and the same categories. The gap between the two — the variance — is what the analysis is actually about.
How to Calculate Variance
The formula is simple:
Variance = Actual − Budget
Often expressed as a percentage too:
Variance % = (Actual − Budget) ÷ Budget × 100
Example: Budgeted marketing spend for the month was $20,000. Actual spend was $24,000.
- Variance = $24,000 − $20,000 = $4,000 over budget
- Variance % = $4,000 ÷ $20,000 × 100 = 20% over budget
That's the mechanical part. The useful part is what comes next.
Favorable vs. Unfavorable Variance — And Why "Favorable" Isn't Always Good News
Variances get labeled favorable or unfavorable based on their direction relative to profit:
- Revenue coming in higher than budgeted → favorable
- Expenses coming in lower than budgeted → favorable
- Revenue coming in lower than budgeted → unfavorable
- Expenses coming in higher than budgeted → unfavorable
The trap: a "favorable" variance isn't automatically good. Expenses coming in under budget because a marketing campaign got cut can show up as a favorable expense variance and an unfavorable revenue variance the following quarter, once the pipeline dries up. Every variance needs a "why," not just a label.
What to Actually Do With a Variance
This is the step most budget vs. actual reports skip entirely. For every meaningful variance:
- Identify the size. Small variances (a few percent) are often just noise. Focus attention on the ones large enough to matter — a common threshold is anything over roughly 10%, or any dollar amount large enough to affect cash flow.
- Find the driver. Was it a timing issue (an expense that shifted into next month), a one-time event, or a real trend? These require completely different responses.
- Decide if the forecast needs to change. A one-time variance usually doesn't change your outlook for the rest of the year. A trend does — and should flow directly into an updated forecast. See Budgeting vs. Forecasting for how the two connect.
- Decide if the business needs to change. Sometimes the right response isn't updating a spreadsheet — it's a real operational decision: cutting a cost, adjusting pricing, revisiting a hiring plan.
A variance report that stops at step 1 is just an accounting exercise. Steps 2 through 4 are where it becomes a management tool.
A Simple Monthly Variance Review Process
For most growing businesses, this doesn't need to be complicated:
- Close the month's books (accurate, reconciled data — this step can't be skipped or rushed)
- Generate the budget vs. actual report by category
- Flag variances above your threshold (both dollar amount and percentage)
- For each flagged item, write one sentence on the driver
- Decide: does this change the forecast, require an operational response, or is it just noise?
- Bring the flagged items — not the whole report — to the leadership review
That last point matters. A 40-line variance report handed to a busy owner or CEO gets skimmed, not used. A one-page summary of the five variances that actually matter gets acted on.
FAQs
What's an acceptable variance percentage?
There's no universal number — it depends on the category and the business. A 20% variance on a small discretionary expense line might be irrelevant; a 5% variance on revenue at a $10M company is a much bigger deal in dollar terms. Most businesses set thresholds by category rather than using one flat percentage across the board.
How often should you run this?
Monthly, at minimum, tied to the regular close process. Businesses managing tighter margins or more volatile cash sometimes review key categories (like revenue and payroll) weekly as well.
What's the difference between variance analysis and forecasting?
Variance analysis looks backward — comparing what happened to what was planned. Forecasting looks forward — using that same information, plus current trends, to predict what's coming next. They work together: variance analysis often supplies the evidence that a forecast needs updating.
Who should own this process?
In smaller businesses, often the owner or a fractional CFO. As the business grows, this typically becomes a defined part of the monthly close rhythm, owned by whoever leads FP&A — see What Is FP&A? for how this fits into the broader function.