Key takeaways
- Variance analysis compares actual results with a budget, forecast, or standard and splits each difference into its causes: price, volume, and mix for revenue, and price, quantity, rate, and efficiency for costs.
- In the worked example, revenue misses budget by only $3,500, but that net figure hides a $136,000 favorable volume variance, a $126,000 unfavorable mix variance, and a $13,500 unfavorable price variance.
- Price and rate variances use actual quantities, and volume, mix, quantity, and efficiency variances use budget or standard prices, so the pieces add up to the total variance with nothing left over.
Variance analysis compares actual results with a budget, forecast, or standard cost and breaks each difference into its drivers, such as price, volume, mix, or efficiency. Getting it right gives you commentary leadership can act on, catches pricing and cost problems early, and keeps your forecast tied to actual results. This guide covers every core formula, one company's revenue and cost variances worked in full, a monthly process, thresholds, commentary examples, multi-entity and multi-ERP analysis, and where AI fits.
- Variance: The difference between an actual amount and the budget, forecast, or standard it's measured against
- Favorable (F): A difference that helps profit, such as revenue above plan or costs below plan
- Unfavorable (U): A difference that hurts profit, such as revenue below plan or costs above plan. Some teams say "adverse" instead.
Variance analysis formulas
The core variance analysis formula is actual minus budget. Each type of variance splits that total into a price or rate effect and a quantity or efficiency effect. Most formulas below are written as actual minus plan, and the last column tells you how to read the sign of each one.
| Variance | Formula | Positive result is |
|---|---|---|
| Budget variance ($) | Actual − Budget | Favorable for revenue, unfavorable for costs |
| Budget variance (%) | (Actual − Budget) ÷ |Budget| × 100 | Same as the dollar variance |
| Sales price variance | (AP − BP) × AQ | Favorable |
| Sales volume (quantity) variance | (Total AQ − Total BQ) × Budget average price | Favorable |
| Sales mix variance | Sum by product of (AQ − Total AQ × Budget mix %) × BP | Favorable |
| Material price variance | (AP − SP) × AQ purchased | Unfavorable |
| Material quantity (usage) variance | (AQ used − SQ) × SP | Unfavorable |
| Labor rate variance | (AR − SR) × AH | Unfavorable |
| Labor efficiency variance | (AH − SH) × SR | Unfavorable |
| Variable overhead spending variance | Actual variable overhead − (AH × Standard variable overhead rate) | Unfavorable |
| Variable overhead efficiency variance | (AH − SH) × Standard variable overhead rate | Unfavorable |
| Fixed overhead spending (budget) variance | Actual fixed overhead − Budgeted fixed overhead | Unfavorable |
| Fixed overhead volume variance | Budgeted fixed overhead − (SH × Standard fixed overhead rate) | Unfavorable |
Key: AP and BP are actual and budget price. AQ and BQ are actual and budget quantity. SP and SR are standard price and standard rate. AR is the actual labor rate and AH is actual hours. SQ and SH are the standard quantity and standard hours allowed for the output you actually produced.
The material, labor, and variable overhead formulas match OpenStax's Principles of Accounting, Volume 2: Managerial Accounting (February 14, 2019), chapter 8. The fixed overhead formulas match the Business LibreTexts Managerial Accounting text, section 10.9: spending is actual minus budgeted fixed costs, and production volume is budgeted minus applied fixed costs. OpenStax computes the material price variance on the quantity used. Many companies compute it on the quantity purchased, so the variance shows up when the material is bought. The cost example below assumes the company used what it bought, so both give the same answer.
Two rules keep the formulas consistent. Price and rate variances multiply by the actual quantity, and volume, mix, quantity, and efficiency variances multiply by the budget or standard price. Follow both and the components add up exactly to the total variance. Terms also vary. The Association for Financial Professionals (AFP) glossary, updated June 29, 2026, calls the volume piece the sales-quantity variance. It describes a multiproduct sales volume variance that combines the quantity and mix effects. OpenStax calls the labor efficiency variance the direct labor time variance.
Budget vs actual and forecast vs actual variance formula
The budget vs actual variance formula is actual minus budget, and the percentage variance divides that difference by the budget. Forecast vs actual works the same way, with the latest forecast in place of the budget.
- Dollar variance: Actual − Budget, or Actual − Forecast
- Percentage variance: (Actual − Budget) ÷ |Budget| × 100, or (Actual − Forecast) ÷ |Forecast| × 100
Example: The Pro line in the worked example below budgeted $600,000 of October revenue, and the September forecast had it at $560,000. Actual revenue was $553,500.
| Comparison | Plan | Actual | Dollar variance | Percentage variance |
|---|---|---|---|---|
| Budget vs actual | $600,000 | $553,500 | −$46,500 (U) | −7.75% |
| Forecast vs actual | $560,000 | $553,500 | −$6,500 (U) | −1.16% |
The forecast had already picked up most of the miss, which is why it helps to run both comparisons.
Is variance actual minus budget or budget minus actual?
Both conventions are in use. This page uses actual minus budget, so a positive number means actual came in above plan. Some cost reports use budget minus actual, so a positive number means spending came in under budget. Either works if you state the convention and label every variance F or U, so no reader has to work out what the sign means.
Divide by the absolute value of the budget. A line budgeted as a $100,000 operating loss that comes in at an $80,000 loss then shows a 20% favorable variance, not a negative percentage that reads as a miss.
A 20% variance means actual landed 20% above or below plan. On a $50,000 cost budget, $60,000 of actual spend is a 20% unfavorable variance.
What is a favorable vs unfavorable variance?
A favorable variance is one that increases profit compared with plan, and an unfavorable variance is one that reduces it. Revenue above budget and costs below budget are favorable, while revenue below budget and costs above budget are unfavorable.
Favorable doesn't always mean good. A labor cost below budget can mean a role is still unfilled. A materials saving can come from a cheaper input that causes quality problems later. The AFP glossary says significant favorable variances should be analyzed too, to see whether they reflect practices worth adopting, lead to quality problems, or cause related unfavorable variances.
What are the types of variance analysis?
The types of variance analysis are grouped by what you compare against, such as budget, forecast, or flexible budget, and by line: price, volume, and mix for revenue, and material, labor, and overhead for costs. Knowing which comparison you're running tells you what the variance can and can't explain.
| Type | What it compares | Use it to |
|---|---|---|
| Budget-to-actual | Actual results with the annual budget | Measure performance against the plan leadership approved |
| Forecast-to-actual | Actual results with the latest forecast | Test how accurate recent forecasts were |
| Forecast-over-forecast | One forecast version with the next for the same months | See how expectations changed and whether a miss is expected to recover |
| Flexible budget | Actual costs with the budget restated at actual volume | Separate spending and efficiency from the effect of selling more or less |
The first three come from Corporate Finance Institute's April 17, 2025 article on the variance types FP&A analysts use. Within each comparison, you break revenue into price, volume, and mix. Variance analysis in cost accounting breaks cost of goods sold into the material, labor, and overhead variances in the formulas table.
Variance analysis example: Price, volume, and mix
In this example, a price-volume-mix analysis shows that a $3,500 revenue miss is the net of a $136,000 favorable volume variance, a $126,000 unfavorable mix variance, and a $13,500 unfavorable price variance.
Example: A company with about $20 million in annual revenue sells three product lines: Standard, Pro, and Enterprise. These are its October 2026 budget and actual results. All amounts are illustrative.
| Product | Budget units | Budget price | Budget revenue | Actual units | Actual price | Actual revenue | Variance |
|---|---|---|---|---|---|---|---|
| Standard | 3,000 | $200 | $600,000 | 3,600 | $190 | $684,000 | $84,000 F |
| Pro | 1,500 | $400 | $600,000 | 1,350 | $410 | $553,500 | $46,500 U |
| Enterprise | 500 | $1,000 | $500,000 | 450 | $1,020 | $459,000 | $41,000 U |
| Total | 5,000 | $340 avg. | $1,700,000 | 5,400 | $314.17 avg. | $1,696,500 | $3,500 U |
A budget vs actual report would show revenue 0.2% under budget and stop there. Units actually beat budget by 8%, and the analysis below shows where the money went.
How to calculate price, volume, and mix variance
Working through the three effects in order gives you a bridge where every dollar of the revenue variance has a named cause.
- Find the budget mix and average price: Standard is 3,000 of 5,000 budget units (60%), Pro is 30%, and Enterprise is 10%. The budget average price is $1,700,000 ÷ 5,000 = $340.
- Calculate the volume variance: (5,400 − 5,000) × $340 = $136,000 favorable. This is what 400 extra units would have earned at the budget mix and budget prices.
- Calculate the mix variance: Restate actual units at the budget mix: 5,400 × 60%, 30%, and 10% = 3,240 Standard, 1,620 Pro, and 540 Enterprise. Then multiply each product's difference by its budget price: (3,600 − 3,240) × $200 + (1,350 − 1,620) × $400 + (450 − 540) × $1,000 = $72,000 − $108,000 − $90,000 = $126,000 unfavorable.
- Calculate the price variance: ($190 − $200) × 3,600 + ($410 − $400) × 1,350 + ($1,020 − $1,000) × 450 = −$36,000 + $13,500 + $9,000 = $13,500 unfavorable.
- Check the total: $136,000 − $126,000 − $13,500 = −$3,500, which matches the revenue variance exactly.
| Product | Volume variance | Mix variance | Price variance | Total variance |
|---|---|---|---|---|
| Standard | $48,000 F | $72,000 F | $36,000 U | $84,000 F |
| Pro | $48,000 F | $108,000 U | $13,500 F | $46,500 U |
| Enterprise | $40,000 F | $90,000 U | $9,000 F | $41,000 U |
| Total | $136,000 F | $126,000 U | $13,500 U | $3,500 U |
The same result as a revenue bridge:
| Step | Revenue | Change |
|---|---|---|
| Budget revenue (5,000 units at budget mix and prices) | $1,700,000 | |
| Volume: 5,400 units at budget mix and prices | $1,836,000 | +$136,000 |
| Mix: 5,400 units at actual mix and budget prices | $1,710,000 | −$126,000 |
| Price: 5,400 units at actual mix and actual prices | $1,696,500 | −$13,500 |
What the analysis tells you:
- Volume: Demand was strong. The company sold 400 more units than planned.
- Mix: Two-thirds of units were Standard against a 60% plan, which pulled the average budget price per unit from $340 to $316.67 ($1,710,000 ÷ 5,400). Across 5,400 units, that drop is the $126,000 mix variance ($1,836,000 − $1,710,000).
- Price: A $10 Standard discount cost $36,000, and higher Pro and Enterprise prices recovered $22,500 of it.
Conventions differ between teams. FTI Consulting's May 3, 2024 white paper on price-volume-mix reverses the order used here. It calculates the price effect on prior-period volume and the volume effect at current prices. Either method reconciles, but the split between price and volume changes with the order, so pick one convention and use it every month.
Cost variance example: Material, labor, and overhead
In the same example, the Standard line's production costs are $15,900 favorable against standard. The flexible budget shows that only $2,100 of the gap with budget comes from spending and efficiency.
Example: The company built 3,600 Standard units in October 2026 against a budget of 3,000. Its standard cost card per unit:
| Cost element | Standard quantity | Standard price | Standard cost per unit |
|---|---|---|---|
| Direct materials (steel) | 4 lbs | $6.00 per lb | $24 |
| Direct labor | 1.5 hours | $30.00 per hour | $45 |
| Variable overhead | 1.5 labor hours | $10.00 per hour | $15 |
| Fixed overhead | $90,000 a month budget | $20.00 per standard hour ($90,000 ÷ 4,500 budgeted hours) | $30 |
Actual October results were 15,000 lbs of steel at $6.20 ($93,000) and 5,200 labor hours at $29.50 ($153,400). Variable overhead was $54,600 and fixed overhead was $93,500. For 3,600 units, the standard allows 14,400 lbs and 5,400 hours.
| Variance | Calculation | Result |
|---|---|---|
| Material price | ($6.20 − $6.00) × 15,000 lbs | $3,000 U |
| Material quantity | (15,000 − 14,400 lbs) × $6.00 | $3,600 U |
| Labor rate | ($29.50 − $30.00) × 5,200 hours | $2,600 F |
| Labor efficiency | (5,200 − 5,400 hours) × $30.00 | $6,000 F |
| Variable overhead spending | $54,600 − (5,200 × $10.00) | $2,600 U |
| Variable overhead efficiency | (5,200 − 5,400 hours) × $10.00 | $2,000 F |
| Fixed overhead spending | $93,500 − $90,000 | $3,500 U |
| Fixed overhead volume | $90,000 − (5,400 × $20.00) | $18,000 F |
| Total | $394,500 actual − $410,400 standard cost of 3,600 units | $15,900 F |
The fixed overhead volume variance is favorable only because the plant built more units than planned, so more fixed cost was absorbed into product cost. It doesn't change what the company spent, so report it separately from the spending and efficiency variances.
Why a flexible budget changes the answer
A flexible budget restates the budget at actual volume, which keeps you from reading higher volume as overspending. Here's the same Standard line compared three ways:
| Comparison | Cost | Variance against actual |
|---|---|---|
| Static budget: 3,000 units × $84 variable cost + $90,000 fixed | $342,000 | $52,500 U |
| Flexible budget: 3,600 units × $84 variable cost + $90,000 fixed | $392,400 | $2,100 U |
| Actual cost | $394,500 |
The static budget makes the line look $52,500 over budget. The flexible budget shows that $50,400 of that is the cost of building 600 more units. The other $2,100 is the net of the seven spending and efficiency variances above. By element, that's $6,600 U for materials, $8,600 F for labor, $600 U for variable overhead, and $3,500 U for fixed overhead spending.
How to do variance analysis, step by step
You do variance analysis each month by calculating variances on closed numbers, breaking the large ones into drivers, confirming causes with owners, and writing commentary that reconciles to the total. A fixed sequence keeps the analysis consistent from month to month and from one analyst to the next.
- Start from closed numbers: Run the analysis on the locked ledger after the close, not a preliminary trial balance, so commentary doesn't change after it's sent
- Match the comparison: Pull the budget and the latest forecast for the same entities, departments, and accounts, and restate cost budgets at actual volume where you use standard costs
- Calculate the dollar and percentage variance: Do it for every line, month and year to date
- Flag what crosses your thresholds: Apply dollar and percentage limits by line, then add anything on your always-investigate list
- Break each flagged variance into drivers: Use price, volume, and mix for revenue, and price, quantity, rate, and efficiency for costs
- Confirm the cause with the owner: Ask the sales, operations, or department lead what happened, and ask for the document that proves it, such as a price change, a purchase order, or a headcount report
- Classify each driver: Mark it timing, permanent, one-off, or error, because that decides whether the forecast changes
- Write commentary and update the forecast: Make the explanations add up to the full variance, then carry permanent drivers into the next reforecast and the month-end reporting pack your FP&A team sends to leadership
Example: If the books lock on business day 5, the analyst runs steps 1 to 5 on business day 6, gets owner answers on business day 7, and sends the package on business day 8.
Which variances should you investigate?
Investigate a variance when it crosses both a dollar and a percentage threshold, or a larger dollar-only limit, with tighter limits for revenue and cost of goods sold than for small expense lines. The AFP glossary calls this management by exception. Leaders focus on variances above a set materiality threshold, find the root cause, judge how controllable it is, and assign an owner.
There's no standard acceptable budget variance percentage. The right limit depends on how large and how predictable each line is, so treat the limits below as a starting point to tune, not a rule.
Example: Monthly thresholds for the example company, with about $1.7 million in monthly revenue:
| Line | Investigate when the variance exceeds | Applied to October |
|---|---|---|
| Revenue by product line | $20,000 and 3% | All three lines flagged: Standard $84,000 F (14.0%), Pro $46,500 U (7.8%), and Enterprise $41,000 U (8.2%) |
| Cost of goods sold by element | $5,000 and 5% | Materials $6,600 U (7.6%) and labor $8,600 F (5.3%) flagged; variable overhead $600 U (1.1%) and fixed overhead spending $3,500 U (3.9%) not flagged |
| Payroll by department | $10,000 and 5% | Not shown in the example |
| Other operating expenses | $5,000 and 10% | Not shown in the example |
| Any line | $50,000 at any percentage | Only Standard revenue ($84,000 F) exceeds it, and it's already flagged |
Total revenue missed by only 0.2%, so a total-level threshold would have flagged nothing. Setting thresholds by product line is what surfaces the mix shift.
Always investigate these, whatever their size:
- Repeating variances: A line that misses in the same direction for three months, even when each month is below threshold
- Large favorable variances: They can point to a missed accrual or delayed spending
- Offsetting variances: Totals that look on plan but hide large opposite moves underneath, as in the revenue example
How to write variance commentary
Strong variance commentary states the amount, the driver, the cause, whether it's timing or permanent, and the forecast impact, and it adds up to the full variance. Weak commentary restates the number or names a cause without sizing it.
Example: Commentary for the example company's October variances.
| Line | Weak commentary | Strong commentary |
|---|---|---|
| Revenue, $3,500 U | "Revenue was slightly under budget." | "Revenue was $3,500 (0.2%) under budget. Units beat plan by 400 (+$136,000), but a shift toward Standard cost $126,000 in mix, and a $10 Standard discount cost $36,000, partly offset by $22,500 of higher Pro and Enterprise pricing. The discount runs through December and is in the forecast." |
| Materials, $6,600 U | "Materials were over budget due to supplier prices." | "Standard materials were $6,600 unfavorable: $3,000 from steel at $6.20 a pound against a $6.00 standard, and $3,600 from using 600 pounds over standard. Purchasing expects the higher price through the first quarter, so it's in the forecast, and production is reviewing scrap." |
| Labor, $8,600 F | "Labor was favorable." | "Standard labor was $8,600 favorable: $6,000 from using 200 fewer hours than standard for 3,600 units, and $2,600 from an average rate $0.50 below standard because two new hires started below the standard rate. The rate effect is permanent; the efficiency gain needs another month before it goes into the forecast." |
Check each comment before it goes out:
- It reconciles: The amounts in the explanation add up to the variance, and if they don't, part of the cause is missing
- It names the driver: Price, volume, mix, rate, or efficiency, with the amount for each
- It says what's next: Timing or permanent, and whether the forecast changes
- It's specific: Words like "various," "timing," or "other" don't appear without an amount and a reason
How to run variance analysis across multiple entities and ERPs
To run variance analysis across entities on different ERPs, map every entity to one reporting chart of accounts, lock every entity on the same close calendar, eliminate intercompany, and split out currency effects before you explain anything else. Skip those steps and part of each variance is a mapping, timing, or exchange rate difference rather than a business driver.
- Map to one reporting chart: Map each ERP's accounts, departments, and product codes to the reporting lines the budget uses, and review the map whenever an entity adds accounts
- Close on one calendar: Start the analysis only when every entity has locked the same period, so one late subsidiary doesn't show up as a variance
- Eliminate intercompany first: Remove intercompany sales, charges, and balances on both sides before you compare with a consolidated budget, as covered in intercompany accounting
- Separate currency effects: Restate actuals at the budget exchange rate to get the constant-currency variance, and report the rest as a foreign exchange variance
- Analyze by entity, then consolidate: Run price, volume, and mix by entity, then check that the entity variances add up to the consolidated variance
Example: A group's UK subsidiary budgets £400,000 of October revenue at a budget rate of $1.30, or $520,000. Actual revenue is £410,000, translated at an actual average rate of $1.25, or $512,500.
| Variance | Calculation | Result |
|---|---|---|
| Constant-currency variance | (£410,000 − £400,000) × $1.30 | $13,000 F |
| Foreign exchange variance | ($1.25 − $1.30) × £410,000 | $20,500 U |
| Total variance in dollars | $512,500 − $520,000 | $7,500 U |
In local currency, the UK business beat its budget by 2.5%. The dollar report shows a 1.4% miss only because the pound came in below the budget rate. Lead the commentary with the operating result and report the currency effect on its own line. For the full sequence across entities, see multi-entity month-end close and month-end consolidation.
Variance analysis vs flux analysis vs budget vs actual
Variance analysis explains gaps against a budget, forecast, or standard, flux analysis explains changes from one period to the next, and budget vs actual is the report that shows the plan gaps.
| Variance analysis | Flux analysis | Budget vs actual | |
|---|---|---|---|
| Compares | Actual with budget, forecast, or standard | Current period with a prior period | Actual with budget, line by line |
| Answers | Which drivers caused the difference, and how much each one contributed | Whether each balance movement is explained and correct | How far each line landed from plan, in dollars and percent |
| Output | Price, volume, mix, rate, and efficiency variances with commentary | Explanations for every movement above threshold | A report with dollar and percentage variances |
| When it runs | After the books lock, for the monthly reporting package | During the close, before the books lock | After the books lock, as the first view of results |
A budget vs actual report gives you the dollar and percentage gap on each line, and this page's formulas break that gap into its drivers. A flux analysis reviews period-over-period changes in balances during the close, so the numbers you analyze against plan are already right.
Can AI do variance analysis?
AI can calculate variances, break them into drivers, and draft commentary, but a person still needs to confirm each cause, decide what changes the forecast, and sign off before the commentary goes out. The arithmetic is the easy part. The judgment about what happened and what it means for next month stays with your team.
| Step | What AI can do | What a person should still review |
|---|---|---|
| Calculate variances | Compute dollar and percentage variances on every line, month and year to date | That the budget version and the account mapping are the right ones |
| Find drivers | Split flagged variances into price, volume, mix, rate, and efficiency, and pull the invoices, orders, or payroll records behind each one | That the split follows your team's convention, so price and volume don't shift between months |
| Explain causes | Draft an explanation from the source documents and the owner's answer | That the cause is real. A draft can name a plausible reason the documents don't support. |
| Classify and forecast | Suggest whether a driver looks like timing, a one-off, or permanent | Whether the forecast changes, which is a call for the FP&A lead |
| Write commentary | Draft commentary that adds up to the full variance | The final wording, and sign-off before it goes to leadership |
Give an AI tool the same inputs an analyst needs: closed numbers, the budget and forecast versions, your thresholds, and access to the source documents. Then test its commentary against the same checklist you'd use for an analyst's draft. For how far AI can go on the close itself, see can AI do the close.
What are the limitations of variance analysis?
Variance analysis is only as timely as your close and only as useful as the standards you measure against. AccountingTools, in a March 3, 2026 article, names three limitations:
- Time delay: Variances compiled at month-end can arrive too late for fast-moving decisions
- Source information: The causes often sit outside the ledger, in bills of material, labor routings, or overtime records, so finding them takes extra work
- Standard setting: A standard that was negotiated or set arbitrarily produces variances that don't tell you much
You reduce all three by reviewing standards at least once a year, keeping operational data next to the ledger, and analyzing only the variances your team can act on.
How we approach it
Every variance in this guide starts from closed numbers. Our Close Agent works your month-end close continuously: it reconciles accounts, investigates exceptions, and prepares adjustments, and every task goes to a named reviewer before the month locks. It works with NetSuite, Oracle, SAP S/4HANA, Microsoft Dynamics 365, Business Central, and QuickBooks, and reads from Odoo, so every entity's books are reconciled and reviewed before your variance analysis starts, whichever ERP they run on. If you want locked numbers sooner each month, explore the Close Agent or book a working session.
Share this article