Variance Analysis (2026): How to Explain Why Every Number Moved

TL;DR

Variance analysis compares actual results with a budget or forecast and explains the gap. The arithmetic is simple: variance equals actual minus budget. The hard part of a monthly review is the explanation, because the reason a number moved usually sits in a supplier email, a pricing call or a promotion agreed in chat rather than in the ledger. This guide covers the formulas, how to split a revenue gap into price, volume and mix, how to set materiality thresholds, a worked revenue bridge and a monthly process. It is published by Sentra, which understands how a company really runs, and acts on it before anyone asks. Sentra joins ERP and BI figures with the conversations behind them, so each variance comment can cite its cause.

What variance analysis measures

Variance analysis compares actual financial results with a budget or forecast to identify differences and investigate their causes. The basic calculation is variance = actual minus budget, as the Association for Financial Professionals glossary sets out. The sign tells you which number is higher. It does not tell you whether the result helped the business, so finance teams label each variance favorable or unfavorable.

Revenue of $110,000 against a $100,000 budget is a favorable $10,000 variance. An expense of $110,000 against the same budget is an unfavorable $10,000 variance. A favorable result still needs scrutiny. Lower labor cost might reflect unfilled roles rather than a lasting saving.

The budget you compare against also shapes the reading. A static budget keeps its original activity level. A flexible budget recalculates expected results at the actual activity level, which lets you separate a change caused by selling more units from a change in revenue per unit.

The variance analysis formulas

Calculate the total variance first, then split it into drivers. For a single product, the price and volume formulas below add back to the total revenue variance.

Total revenue variance = actual revenue - budget revenue

Price variance  = (actual unit price - budget unit price) x actual units sold
Volume variance = (actual units sold - budget units) x budget unit price

Price variance + volume variance = total revenue variance

Price uses actual units because it measures the effect of the price change on everything you sold. Volume uses budget price because it values the change in units without counting the price change a second time.

With several products, volume hides two movements. You may sell more units overall, or customers may buy a different share of each product. To report mix separately, split volume into quantity and mix instead of adding a mix figure on top of the unsplit volume variance.

Quantity variance = (actual total units - budget total units) x budget weighted average unit price
Mix variance (per product) = (actual mix % - budget mix %) x actual total units x product budget unit price

Quantity variance + sum of mix variances + sum of price variances = actual revenue - budget revenue

Each product's mix percentage is its share of total units. If you add a mix variance to the original product-level volume variance, you count the mix effect twice. Use either the plain volume calculation or the quantity and mix split, not both.

Setting materiality thresholds

Set thresholds before the close so the team investigates meaningful exceptions rather than every difference. The AFP calls this management by exception: finance reviews variances above a predetermined threshold, then finds their cause and owner.

A common design uses a dollar floor and a percentage test together, plus a higher absolute override. For example, investigate any variance that exceeds both $50,000 and 5% of budget, and any variance above $500,000 regardless of percentage. A $60,000 miss on a $2 million line is 3% and fails the percentage test. A $510,000 miss on a $20 million line is 2.55% but trips the $500,000 override. Apply the same tests to favorable variances, since lower spending can reflect delayed work.

Regulatory standards give context rather than a universal cutoff. PCAOB AS 2305 asks auditors to set the difference they will accept before comparing expected and recorded amounts, and SEC Regulation S-K Item 303 requires covered companies to discuss material changes in results. Neither sets your monthly investigation threshold.

Worked example: a budget vs actual revenue bridge

This illustrative example reconciles a $12,900 shortfall across volume, product mix, channel mix, price and timing without counting any change twice. The budget assumed 1,000 units and $146,000 of revenue: 500 units each of products A and B, with 20% of each sold through a partner channel. Direct prices were $100 for A and $200 for B, and partner prices were $20 lower for either product.

Actual recognized revenue was $133,100, an unfavorable variance of $12,900. The company billed 900 units, 360 of A and 540 of B. The partner channel took 270 units, or 30% of each product. Every unit sold for $5 less than its budgeted channel price, and $1,000 of billed revenue was deferred to the next month. Change one assumption at a time to build the bridge.

The bridge shows where revenue moved. It cannot tell you why customers bought fewer units, why more volume went through partners or who approved the $5 price cut. Those answers need the records and conversations around the numbers, and a structured root cause analysis template helps when a variance is large enough to warrant a full investigation.

The monthly variance analysis process

• Lock the comparison base. Record the approved budget or forecast version, the period, the currency and the reporting scope so reviewers can reproduce the comparison.
• Check the inputs. Reconcile actuals to the ledger and confirm that budget and actual use the same account mappings. Separate late postings and reclassifications from real changes in activity.
• Calculate and decompose the gaps. Calculate each amount and percentage variance, then split material revenue gaps into price, volume and mix. Confirm the drivers add back to the total.
• Apply the thresholds. Use the amount floor and the percentage test to select variances for review. Also review smaller gaps that recur or signal a risk the thresholds would miss.
• Verify the cause with an owner. Ask the person responsible for the result what changed, then check the answer against purchase orders, pricing approvals or other records.
• Write the commentary and get sign-off. Tie each explanation to a driver, an amount, an owner and a next action. The owner confirms the cause and the finance lead approves the commentary before it goes to leadership.

A consistent sentence pattern keeps commentary checkable. Copy the template below for each material variance. If the cause is still unconfirmed, say so and name the person who will verify it.

[Evidence and source] confirms that [driver] caused a [amount] [favorable / unfavorable]
variance against [budget version] for [period].
[Owner] will [action] by [date].
Status: [confirmed / unconfirmed, verifier and date]

Why the arithmetic is the easy part

A budget vs actual report shows what moved. A variance calculation can show how much of a revenue miss came from price or volume, but it cannot establish why customers paid less or ordered fewer units. Two months can carry the same price variance for different reasons, such as an approved promotion in one month and an unplanned discount in the other.

The cause often sits outside the financial data. An ERP records the invoice price, while the reason for that price may sit in a sales call or a promotion agreed in chat. A supplier email may explain a cost increase that BI shows only as a higher purchase price. A useful explanation connects the recorded transaction to a dated decision or event. If a salesperson cites a promotion, check that its dates, eligible customers and discount match the invoices behind the variance. Without that check, the commentary offers a plausible story rather than an evidenced cause, and the decision is lost by next quarter unless someone writes it into a decision log.

Common mistakes that break variance analysis

• Using the wrong budget version. Compare actuals with the version approved for the review and note any assumptions that have since changed. Do not revise the original budget to make a predictable cost increase disappear.
• Ignoring external changes. Check exchange rates, regulation and market prices before assigning a variance to an internal decision.
• Treating a one-time expense as recurring. Separate an unusual charge from costs likely to continue, or the forecast will carry it into future months.
• Overlooking seasonality. Compare seasonal activity with the planned month, not an annual average. A holiday peak can look favorable even when sales missed the seasonal budget.
• Counting the same gap twice. Use one consistent price, volume and mix method and check that the components sum to the total.
• Losing the trail behind an explanation. Record the source of each figure and the evidence for its cause. A spreadsheet total cannot show whether a supplier email supports the commentary.

How Sentra explains the why behind the numbers

Sentra understands how a company really runs, and acts on it before anyone asks. It reads email, chat, meetings and documents alongside ERP, CRM, HRIS and BI data, and works out what is true now and why. If freight expense runs over budget, Sentra can connect the variance to the supplier email announcing a rate increase, or to the meeting where operations approved a different shipping plan, and raise the developing gap to the person who can act before the close.

A named person approves anything Sentra raises before it reaches anyone else. Every claim cites its source, and access follows the permissions of that source, so a reviewer sees only what they could already open. Sentra replaces none of the company's systems. The ledger, the BI layer and the planning model stay where they are. In Sentra's own measurement at Lenskart as of October 2026, time to root cause fell from 3 to 4 days to about 30 minutes.

FAQ