Variance analysis: budget vs actual in 5 steps

A variance analysis compares actual figures with the budget or forecast and explains the differences. Calculating them is not the work; a spreadsheet does that in a second. The work is deciding which differences matter, where they come from and what you are going to do about them.

Every month brings the same moment: the figures are in, the budget sits beside them, and the board wants to know why the result is off. A column of differences is not an answer yet. Below you will find how to build a variance analysis in five steps, how to split a revenue variance into a price and a volume variance, and which mistakes are worth avoiding.

What is a variance analysis?

A variance analysis sets actuals against a benchmark and explains the gap. A useful analysis answers three questions:

1. How big is the difference? In currency and in percent.
2. What caused it? Price, volume, timing, a one-off item or an error in the budget.
3. What do we do about it? Correct course, accept it or update the forecast.

Budget vs actual or forecast vs actual?

Comparing with the budget measures how far you are from the original plan. Comparing with the forecast measures how well you judged your latest expectation. For the board, budget vs actual is usually the starting point, because the budget is the agreement. For steering during the year the forecast is often more relevant, because known variances are already built in. The distinction is set out in planning, budgeting and forecasting.

Variance analysis in 5 steps

Step 1: Make actuals and budget comparable

Most variances that are not really variances start here. Before you analyse, check:

• The same structure. Actuals and budget must sit on the same general ledger accounts or reporting groups. A budget on "total personnel costs" against actuals on twelve separate accounts does not work.
• The same period. Choose deliberately between month and year-to-date. Monthly figures are sensitive to timing; YTD smooths that out.
• A closed month. Are all invoices booked, depreciation processed and accruals included? A missing journal entry creates a phantom variance. See also month-end closing.
• For a group: consolidated. Intercompany revenue and costs must be eliminated, or you are comparing inflated figures. How that works is covered in eliminating intercompany transactions.

Step 2: Calculate the variance in currency and percent

The formula is simple:

Variance (€) = actual − budget
Variance (%) = (actual − budget) / budget × 100

Always look at both. A 40% variance on a € 2,000 line matters less than 3% on € 2 million of revenue.

Step 3: Label favourable and unfavourable, not plus and minus

A positive number is not always good news. Revenue above budget is favourable, costs above budget are unfavourable, yet both produce a positive difference. Put a label next to every variance: favourable (F) or unfavourable (U). It prevents the classic confusion in the management meeting.

Step 4: Filter with a materiality threshold

Not every difference deserves commentary. Agree a threshold, for example: explain any variance larger than € 10,000 or 10% of budget. Scale the threshold to the size of your organisation. The point is to focus attention on the three or four lines that drive the result, not on twenty small differences.

Step 5: Explain the cause and attach an action

For every material variance, write down briefly:

• Cause: what happened?
• Type: structural, one-off or timing?
• Action: what do we do, and does the forecast need to change?

Timing differences, such as a cost that lands a month later, resolve themselves. Structural differences, such as a higher purchase price that sticks, belong in the forecast.

Worked example: variance analysis of a profit and loss statement

A trading company compares its first-quarter actuals (YTD) with budget. The threshold is € 10,000 or 10%. The figures are illustrative; the structure is the common one.

Line (YTD)	Budget	Actual	Variance	%	F/U	Explain?
Revenue	€ 500,000	€ 487,600	−€ 12,400	−2.5%	U	Yes
Cost of goods sold	€ 300,000	€ 285,200	−€ 14,800	−4.9%	F	Yes
Gross margin	€ 200,000	€ 202,400	+€ 2,400	+1.2%	F
Personnel costs	€ 120,000	€ 131,500	+€ 11,500	+9.6%	U	Yes
Other operating costs	€ 45,000	€ 43,800	−€ 1,200	−2.7%	F	No
EBITDA	€ 35,000	€ 27,100	−€ 7,900	−22.6%	U

At first glance lower revenue looks like the problem. But gross margin is actually above budget. The € 7,900 shortfall in EBITDA comes almost entirely from personnel costs.

Splitting the revenue variance into price and volume

The € 12,400 revenue variance only means something once you split it. The budget assumed 10,000 units at € 50. Actual was 9,200 units at € 53.

Volume variance = (actual volume − budget volume) × budget price
(9,200 − 10,000) × € 50 = −€ 40,000 (U)

Price variance = (actual price − budget price) × actual volume
(€ 53 − € 50) × 9,200 = +€ 27,600 (F)

Together: −€ 40,000 + € 27,600 = −€ 12,400, which ties back to the table.

The story is now quite different. The company sold 8% fewer units but achieved a higher price. That raises a better question: is the lower volume a consequence of the price increase?

The same split for purchasing

The budget assumed a purchase price of € 30 per unit; the actual price was € 31.

• Volume effect: (9,200 − 10,000) × € 30 = € 24,000 lower costs (F)
• Price effect: (€ 31 − € 30) × 9,200 = € 9,200 higher costs (U)

Net € 14,800 favourable. Most of the "saving" on purchasing is therefore not a saving at all: fewer units were sold. The purchase price itself is 3.3% above budget, and that is structural.

The commentary for the board

The same figures produce three lines of commentary:

1. Revenue −€ 12,400 (U): 800 fewer units sold (−€ 40,000), partly offset by a higher selling price (+€ 27,600). Action: review the effect of the price increase on volume with sales.
2. Purchase price +€ 9,200 (U): the supplier raised prices from February. Structural; include in the forecast.
3. Personnel costs +€ 11,500 (U): two temporary staff hired to cover sick leave. One-off, expected until the end of April.

That is a variance analysis a board can act on. Where that commentary fits in the rest of the pack is covered in management reporting.

Common mistakes in variance analysis

• A budget that is not phased. An annual budget divided by twelve ignores seasonality, holiday pay and one-off items. Every month then shows phantom variances.
• Explaining everything. Twenty one-line explanations are worth less than three good ones.
• Variances without a cause. "Personnel costs are above budget" is a description, not an explanation.
• Mixing up timing and structural. Treat a structural price increase as timing and the forecast will trail reality for the rest of the year.
• Copying between files by hand. Every time actuals are pasted from the accounting system into an Excel model, there is a chance of a wrong period or a missed account. With several entities that risk multiplies.

Variance analysis in Smartbooks

In Smartbooks, actuals, budget and forecast sit in the same model, connected to the accounting system. In the report configurator you add variance columns, absolute or in percent, and choose which categories to compare: actual against budget, or actual against forecast, per period or YTD. If you work with several entities, you budget per entity and the budgets roll up automatically to the consolidated level. Projections also show what the variance so far means for the expected full-year result.

Frequently asked questions

What is the difference between variance analysis and budget variance analysis?

Budget variance analysis is variance analysis with the budget as the benchmark. The same method works against a forecast or against last year; only the reference changes.

How often should you run a variance analysis?

Most mid-sized companies run a variance analysis monthly, straight after the month-end close. Smaller organisations sometimes choose quarterly. The sooner after the period, the better your chance to still correct course.

What materiality threshold is common?

There is no fixed rule. A common approach combines an absolute amount with a percentage, for example € 10,000 or 10%. The amount stops small lines with large percentage swings from taking all the attention; the percentage catches relatively large shifts.

How do you calculate a price variance and a volume variance?

Volume variance = (actual volume − budget volume) × budget price. Price variance = (actual price − budget price) × actual volume. Together they equal the total revenue variance.

Should I change the budget when there is a large variance?

No, the budget usually stays as a fixed benchmark. Structural variances go into the forecast. That way you see both how far you are from the original plan and what you now realistically expect.

A variance analysis only adds value once it goes beyond a column of differences. Make the figures comparable, filter with a threshold and split large variances into their causes. To see how budget and actuals line up straight from the ledger, have a look at budgeting or reporting. More articles for finance professionals sit in the knowledge center.