Budget vs actual variance analysis that survives a board meeting
Build a budget vs actual variance analysis that holds up: materiality thresholds, price/volume/mix drivers, and timing versus permanent.
Budget vs actual variance analysis compares each line to plan, then explains the gap by driver. Variance = actual − budget, favourable when profit is higher. Set a materiality rule first — for example explain anything above $50,000 or 10% of the line — then give one sentence per driver, separating timing differences from permanent ones.
Key takeaways
- Set the materiality rule before you look at the numbers: a dollar floor OR a percentage of the line, whichever triggers first. In the example below, five lines cover 98.5% of the total absolute variance.
- Every variance gets one sentence: driver, dollar amount, cause, action, owner. If it takes a paragraph, you have not found the driver yet.
- Split the revenue variance into price, volume and mix. "Revenue is down $504,000" is not an explanation; "volume −$400,000, mix −$95,000, price −$9,000" is.
- Label every variance permanent or timing. A quarter that looks $285,000 light can be $416,000 light underneath once timing benefits are removed.
- A favourable variance is not automatically good news. Favourable COGS driven by lower volume, or favourable marketing driven by a campaign that did not run, are both bad news wearing the wrong sign.
Most variance packs fail in the same way. They list every line with a difference, colour the negatives red, and attach commentary that repeats the number in words: "Marketing is $172,000 favourable due to lower spend." That sentence tells the board nothing they could not read off the column. A budget vs actual variance analysis is only worth building if it survives the follow-up question, and the follow-up question is always why, and does it repeat?
- Variance
- The difference between an actual result and the budget or forecast for the same period. A variance is favourable when it increases profit and unfavourable when it reduces it — which means a cost above budget and revenue below budget are both unfavourable, despite having opposite arithmetic signs.
Start with a materiality rule, not the numbers
Decide what deserves explanation before you know which lines moved. Otherwise the threshold gets set retrospectively to exclude whatever is awkward. A workable rule has two legs, because either alone fails:
- A dollar floor, so trivial percentages on large lines are ignored. A 1% miss on a $12,400,000 revenue line is $124,000 and matters; a 40% miss on a $3,000 line does not.
- A percentage of the line, so large percentage moves on small lines are caught. A professional fees line that doubles is a story even if the dollars are modest, because it is often a leading indicator.
- An override for the top five by absolute value, explained regardless of whether they trip the rule. It stops a threshold from being gamed.
For a business of the size below, "explain anything where the variance is at least $50,000 or at least 10% of the budgeted line" is a reasonable rule. Scale the dollar floor to something like 0.5% of quarterly revenue and keep it fixed for the year.
The variance table
A quarter of budget vs actual for a mid-market manufacturer. Signs are shown from the profit perspective: positive means favourable.
| Line | Budget | Actual | Variance | % of line | Explain? |
|---|---|---|---|---|---|
| Revenue | $12,400,000 | $11,896,000 | −$504,000 | 4.1% | Yes — both tests |
| COGS | $7,564,000 | $7,375,000 | +$189,000 | 2.5% | Yes — dollar test |
| Gross profit | $4,836,000 | $4,521,000 | −$315,000 | 6.5% | — |
| Payroll | $2,150,000 | $2,207,000 | −$57,000 | 2.7% | Yes — dollar test |
| Marketing | $620,000 | $448,000 | +$172,000 | 27.7% | Yes — both tests |
| Facilities | $310,000 | $316,000 | −$6,000 | 1.9% | No |
| Professional fees | $145,000 | $233,000 | −$88,000 | 60.7% | Yes — both tests |
| Other opex | $268,000 | $259,000 | +$9,000 | 3.4% | No |
| Operating profit | $1,343,000 | $1,058,000 | −$285,000 | 21.2% | — |
Two things to notice before any commentary is written. First, operating profit is 21.2% below plan while revenue is only 4.1% below — that is operating leverage, and it is the headline. Second, the five explainable lines cover almost all the movement, so the pack needs five sentences, not eighteen.
Decompose the revenue variance
"Revenue is $504,000 below budget" is a restatement, not an explanation. Break it into price, volume and mix using the same additive decomposition as a price volume mix analysis, with budget standing in for the prior period — the price volume mix calculator takes budget and actual units and prices for up to five lines. Two product lines here:
| Line | Budget units | Budget price | Actual units | Actual price |
|---|---|---|---|---|
| Core | 50,000 | $170.00 | 49,000 | $166.00 |
| Pro | 12,000 | $325.00 | 11,000 | $342.00 |
| Total | 62,000 | — | 60,000 | — |
Total unit growth factor g = 60,000 ÷ 62,000 = 0.967742, so units are 3.2% below plan.
- Price = (actual price − budget price) × actual units. Core: (−$4.00) × 49,000 = −$196,000. Pro: (+$17.00) × 11,000 = +$187,000. Total −$9,000.
- Volume = budget revenue × (g − 1) = $12,400,000 × (−1 ÷ 31) = −$400,000. By line: Core −$274,194, Pro −$125,806.
- Mix = budget price × (actual units − budget units × g). Core: 49,000 − 48,387.1 = +612.9 units × $170 = +$104,194. Pro: 11,000 − 11,612.9 = −612.9 units × $325 = −$199,194. Total −$95,000.
- Check: −$9,000 − $400,000 − $95,000 = −$504,000, equal to the revenue variance to the dollar.
Then decompose gross profit
COGS is $189,000 favourable, and that number is a trap. Standard costs are $104.00 for Core and $197.00 for Pro, which reconciles to budget: (50,000 × $104) + (12,000 × $197) = $5,200,000 + $2,364,000 = $7,564,000. Rebuild COGS the same three ways as revenue.
- COGS at actual volumes and standard costs = (49,000 × $104) + (11,000 × $197) = $5,096,000 + $2,167,000 = $7,263,000
- Volume = (60,000 − 62,000) × budget blended cost of $122.00 = −$244,000 of cost, favourable, and entirely a consequence of shipping less
- Mix = $104 × (+612.9) + $197 × (−612.9) = $63,742 − $120,742 = −$57,000 of cost, favourable, because the shift ran toward the cheaper Core line
- Rate = actual $7,375,000 − standard-cost $7,263,000 = +$112,000 of cost, unfavourable: a genuine $1.87 per unit overrun against standard
- Check: −$244,000 − $57,000 + $112,000 = −$189,000 of cost, which is the $189,000 favourable variance
Note what the full split reveals that the blended figure hid. Actual cost per unit was $7,375,000 ÷ 60,000 = $122.92 against a budget blend of $122.00, apparently a 92-cent miss. But mix moved toward the cheaper product, so the blend should have fallen. Measured properly against each product's own standard, the overrun is $1.87 a unit — twice the headline. Blended margin at both levels is worth recomputing from the underlying lines rather than the summary; the gross margin calculator gives margin, markup and margin per unit from one set of inputs.
Now the gross profit variance splits four ways, and the four sum exactly to the reported number:
| Driver | Effect on gross profit | Where it came from |
|---|---|---|
| Price | −$9,000 | Core discounting of $4.00/unit, offset by Pro pricing $17.00 above plan |
| Volume | −$156,000 | 2,000 units short of plan at $78.00 of budgeted gross profit per unit |
| Mix | −$38,000 | 613 units shifted from Pro ($128 profit/unit) to Core ($66 profit/unit) |
| Unit cost vs standard | −$112,000 | $1.87 per unit above standard cost on 60,000 units |
| Total | −$315,000 | Equals the gross profit variance exactly |
That table is the difference between a variance report and a variance analysis. Four numbers, four owners, and the useful insight is now visible: volume is the largest single driver, but the $112,000 cost overrun and the $9,000 of net price are the ones that persist into next quarter if nothing changes.
Cost out the hours your team spends assembling variance packs each month — and what recovering them is worth in FTE terms.
Permanent or timing: label every variance
This is the classification that separates a competent pack from a credible one. A timing difference reverses in a later period; a permanent difference does not. Confusing them causes the specific failure of a favourable quarter followed by a surprise.
| Line | Variance | Permanent | Timing | Why |
|---|---|---|---|---|
| Revenue | −$504,000 | −$504,000 | — | Units and price are realised; nothing reverses |
| COGS | +$189,000 | +$189,000 | — | Volume-driven saving plus a real cost overrun, both realised |
| Payroll | −$57,000 | −$38,000 | −$19,000 | Two roles filled above the planned salary band (permanent run-rate); bonus accrual catch-up (timing) |
| Marketing | +$172,000 | +$22,000 | +$150,000 | Campaign deferred one quarter, not cancelled |
| Professional fees | −$88,000 | −$88,000 | — | One-off legal matter, incurred and closed |
| Facilities / other | +$3,000 | +$3,000 | — | Below materiality, netted |
Restate the result on that basis. Reported operating profit variance is −$285,000. Remove the $150,000 of favourable marketing timing and add back the $19,000 of unfavourable payroll timing: underlying variance is −$416,000. The quarter is materially worse than the headline, and the marketing spend will land next quarter on top of whatever else happens. That single sentence is usually the most valuable line in the pack.
One sentence per driver
The discipline that makes a variance story survive scrutiny: each explanation is one sentence containing the driver, the dollar amount, the cause, the action and the owner. If you need a paragraph, you have not identified the driver — you are describing the symptom. Five sentences for the quarter above:
- 1.Revenue −$504,000 (4.1%): volume 2,000 units below plan (−$400,000), Pro-line mix 613 units light (−$95,000), net price broadly on plan (−$9,000); the volume shortfall is two delayed OEM programmes now scheduled for Q3, owned by sales.
- 2.Gross margin 1.0 point below plan at 38.0%: $112,000 of unit cost above standard ($1.87/unit) and $38,000 of adverse mix; a standard-cost review is scheduled before the next quarter, owned by operations.
- 3.Payroll −$57,000 (2.7%): two engineering roles filled above the planned salary band (−$38,000, a $152,000 annualised run-rate) plus a bonus accrual catch-up (−$19,000 timing); no further hires before Q4, owned by the CTO.
- 4.Marketing +$172,000 (27.7%): the Q2 channel campaign moved to Q3 (+$150,000 timing) with $22,000 of permanent saving on media rates; the deferred spend will land in Q3, owned by marketing.
- 5.Professional fees −$88,000 (60.7%): one supplier dispute resolved through outside counsel; matter closed, no further exposure expected, owned by the general counsel.
Common failure modes
- Restating the number in words. "Revenue is below budget due to lower sales." Every explanation must name a physical event: a lost account, a delayed programme, a price concession, a supplier increase.
- Comparing to the wrong baseline. Original budget, latest forecast and prior year answer different questions. Show budget for accountability and forecast for decisions, and label which is which in the column header.
- Period-alignment errors masquerading as variances. An accrual missed at cut-off produces a favourable variance this month and an unfavourable one next month. Two consecutive equal-and-opposite variances is an accounting signature, not a business event — the month-end close checklist is where these get caught.
- Treating favourable as good. Favourable purchase price variance can mean a cheaper substitute material that raises scrap downstream. Check the purchase price variance calculator alongside the quality and scrap numbers before you claim a win.
- Explaining percentages of percentages. "Gross margin is 2.6% below budget" is ambiguous between 1.0 percentage points and 2.6% of 39.0%. Always say points for margin movements and reserve percent for dollar changes.
- Variance analysis with no action column. A variance with no owner and no next step is commentary. Every material line needs a name against it.
Getting the variance pack built without the week
None of the arithmetic above is hard. The work is assembly: budget loaded at the right dimension, actuals in the same period alignment, revenue split by product and price, cost split between rate and volume, department mapping consistent between the two, and all of it repeated every month. Native budget vs actual reporting gets you the columns; the drivers still come from exports and pivots.
That assembly time is worth costing. Three analysts at six hours a week each on variance packs is 18 hours weekly, or 828 hours a year across 46 working weeks. At a $65 loaded hourly cost that is $53,820 a year, or 0.44 of an FTE, spent moving numbers between systems rather than explaining them.
This is the class of question answers directly. Ask "gross margin by product family, budget versus actual for Q2, with unit volumes" and get the figures computed live from your own account, with the query shown underneath so you can confirm the period and dimension filters before the numbers reach a board pack.
Frequently asked questions
What is a good materiality threshold for variance analysis?
Use two tests joined by OR: a dollar floor around 0.5% of period revenue, and a percentage of the individual line around 10%. Explain anything that trips either, plus the five largest variances by absolute value regardless. Set the rule before you see the results and keep it fixed for the year so the series stays comparable.
Is a favourable variance always good?
No. Favourable COGS caused by lower volume simply reflects a revenue miss. Favourable marketing caused by a campaign that did not run is deferred spend, not a saving. Favourable purchase price variance can come from a cheaper material that raises scrap later. Always identify the driver before recording a favourable variance as good news.
What is the difference between a permanent and a timing variance?
A timing variance reverses in a later period: a deferred campaign, an accrual booked late, an invoice that slipped past cut-off. A permanent variance does not: a lost customer, a price concession, a one-off legal fee. Label every material variance as one or the other, then restate the result excluding timing to show the underlying position.
How do you explain a revenue variance properly?
Split it into price, volume and mix, and make the three sum exactly to the total. Price is the change in realised price times actual units; volume is budget revenue times the overall unit shortfall; mix is each line's units above or below its proportional share at budget prices. Then name the physical event behind the largest driver.
What is flux analysis and how does it differ from variance analysis?
Flux analysis explains the change in a balance or account between two periods — this month against last month, or this year against prior year. Variance analysis explains the gap against a plan. The mechanics are the same; the baseline differs. Most close packs need both: flux for balance sheet movements, variance for P&L accountability.
How many variances should a board pack actually explain?
Enough to cover the great majority of total absolute variance, which is usually four to six lines. In the example above five lines account for 98.5% of $1,025,000 of movement. Listing every line with a difference dilutes attention and invites questions about immaterial items instead of the drivers that matter.
Calculators for this
Reporting time savings calculator — price the hours your team spends building reports, the annual cost, hours recovered and the FTE equivalent.
Split a revenue change into price, volume and mix effects across up to five product lines. The three effects reconcile exactly to the total variance.
Purchase price variance calculator: PPV from standard and actual price times quantity received, with favourable or unfavourable and annual exposure.
Calculate gross margin from revenue and COGS. Get gross profit, margin %, markup %, margin per unit, and the price you need for a target margin.
Keep reading
Price volume mix analysis with a fully worked example: an additive decomposition where price, volume and mix sum exactly to the revenue change.
Build a NetSuite budget vs actual report that reconciles: budget loading, period alignment, variance columns and department dimensions.
An ordered NetSuite month-end close checklist: 14 steps with the period-locking, inventory-costing, revenue and intercompany gotchas that break a close.
Margin vs markup, settled: both formulas, a checked conversion table, and the pricing error that quietly costs you dollars per unit.