Variance Analysis: Formula, Methods, and How to Automate

Why did the numbers miss plan, and can you answer that in one sentence? Variance analysis is how finance teams get to that answer, and doing it well is the difference between a report people read and one they skim.

This guide is for controllers, FP&A analysts, and founders who produce a monthly pack and suspect the commentary is not earning its keep.

We help global companies hire FP&A and reporting analysts in India through our Employer of Record service, so the approach to variance analysis below is the one we see working on real monthly packs.

Most explanations give you the subtraction and stop. We go to the part that actually changes decisions: splitting a variance into price and volume, setting a threshold worth investigating, and writing commentary that names a cause.

We close on who does this work when the finance team is small, which is where our guide to an offshore FP&A team in India picks up in more detail.

What is variance analysis?

Variance analysis is the practice of comparing actual results against a budget, forecast, or prior period and explaining the difference. The output is not the number itself but the reason behind it, stated clearly enough for someone to act on.

The subtraction is trivial. The explanation is the job.

A report saying revenue came in $180,000 below budget has told you nothing you could not see yourself. One saying the miss came entirely from a 6 percent discount on renewals, while volumes held, has told you where to look.

That is the standard to hold your own pack to.

Start with the arithmetic, then we will get to the part that matters.

How do you calculate a variance?

Subtract the budgeted figure from the actual figure. Divide that difference by the budgeted figure to express it as a percentage. Report both, because a large percentage on a small account and a small percentage on a large one need different responses.

The base formulas

Variance = Actual Result - Budgeted Result

Variance Percentage = (Actual Result - Budgeted Result) / Budgeted Result

Favorable and unfavorable

A variance is favorable when it improves profit against plan and unfavorable when it reduces it. Note that the sign flips between revenue and cost lines: revenue above budget is favorable, while cost above budget is not.

It is worth saying plainly that favorable does not mean good. An underspend on hiring looks favorable and may mean you failed to fill roles the plan depended on.

Treat the label as a direction, never as a verdict.

A worked example

Budgeted revenue for the quarter was $50,000, built from 1,000 units at $50 each. Actual revenue came in at $52,900, from 1,150 units at an average $46.

The total variance is $52,900 minus $50,000, or $2,900 favorable. That is 5.8 percent above budget, and on most monthly packs the commentary would stop there and call it a beat.

Stopping there would hide the actual story, which the next section pulls apart.

How do you split a variance into price and volume?

Multiply the price difference by actual volume to get the price variance, then multiply the volume difference by budgeted price to get the volume variance. The two add back exactly to the total, so nothing is left unexplained.

This is the most useful technique in the whole topic, and it is also the easiest one to leave out of a monthly pack.

Price Variance = (Actual Price - Budgeted Price) x Actual Volume

Volume Variance = (Actual Volume - Budgeted Volume) x Budgeted Price

Run those against the worked example above and the favorable headline falls apart in a useful way.

Price variance is ($46 minus $50) multiplied by 1,150 units, which is $4,600 unfavorable. Volume variance is (1,150 minus 1,000) multiplied by $50, which is $7,500 favorable.

Add them together and you get exactly the $2,900 favorable total. The reconciliation is a built-in check on your own work.

Read the table and the real finding is obvious: the business sold 15 percent more units at an 8 percent lower price. Revenue beat plan, and unit economics got worse.

That is a discounting question for the sales leadership, and it would have been invisible in a report that only showed the $2,900 beat.

Building this split into the standard monthly pack is largely a reporting design decision, and we cover how teams structure that in our guide to internal reporting operations.

It also depends on having unit-level data to split, which is often the real blocker rather than the formula. Our guide to offshore data analytics in India covers getting that foundation in place.

Price and volume are the two you will use most, but they are not the only splits available.

What are the main types of variance?

Four appear in most management reporting: price or rate variance, volume or quantity variance, mix variance where the blend of products shifts, and efficiency variance where the inputs consumed differ from standard. Cost accounting subdivides these further.

You do not need all four every month. Pick the ones that map to decisions someone in your business can actually make.

Mix variance is the one teams most often ignore and later wish they had tracked, particularly in businesses with several product lines closing through the same record to report process.

Here is what each type tells you, and who normally needs to hear it.

Notice the last column. A variance nobody owns is a number, not an insight, and it will keep reappearing every month unchanged.

Working capital lines deserve the same treatment, and if collections is where your variances keep appearing, our guide to an offshore accounts receivable team in India picks up that thread.

On the cost side, unexplained spend variances often trace back to invoices nobody approved properly, which is a payables problem. Our guide to an offshore accounts payable team in India covers that side.

Which brings us to how the finding actually gets communicated.

What is flux commentary and why does it matter?

Flux commentary is the written explanation attached to each material variance. Flux is short for fluctuation, and the commentary is what turns a variance table into something a reader can act on. Auditors and boards read it closely.

Most commentary fails in the same way: it restates the number instead of explaining it.

Weak version: "Professional fees were $84,000 over budget in Q3." The reader already knew that from the column next to it.

Strong version: "Professional fees ran $84,000 over budget because the data migration slipped into Q3 and required six extra weeks of external engineering. The work is complete and Q4 returns to plan."

Same variance, completely different value to the reader. Good commentary answers four questions every time:

• What moved: the account, the amount, and the direction, stated once and briefly.
• Why it moved: a specific business event, not a category label like "higher costs".
• Whether it repeats: one-off, timing, or a permanent run-rate change, because the three call for different responses.
• What happens next: the action already taken or the decision now needed, with an owner.

The timing distinction in the third point is the one that saves the most argument later, because a timing variance reverses itself and a run-rate variance does not. Teams that produce this well usually have dedicated reporting analysts drafting it rather than leaving it to whoever closed the ledger.

Of course, nobody has time to write that for every line, which is where thresholds come in.

How do you set a materiality threshold for investigation?

Use a percentage test and a dollar test together, and investigate only variances that breach both. A percentage test alone floods you with trivial swings on small accounts. A dollar test alone lets large percentage moves on small lines slip through.

A common starting point is 5 percent and a dollar floor scaled to the size of your business. The exact numbers matter less than applying them consistently.

Say you set 5 percent and $25,000. A marketing line $12,000 over on a $60,000 budget is 20 percent but under the dollar floor, so it does not qualify. A payroll line $40,000 over on a $2 million budget is 2 percent, so it does not either.

That second one deserves a caveat, because a 2 percent payroll variance can still matter if it signals a permanent headcount change. Build in three overrides:

• Always investigate a new or unexpected account: spend appearing where the budget had none is worth a look at any size.
• Always investigate a persistent small variance: the same line missing by 3 percent for six months is a broken assumption, not noise.
• Always investigate anything affecting a covenant or a board metric: materiality is about consequence, not size.

Set those rules once, write them into the close checklist, and the monthly argument about what to explain largely disappears. It is one of the quiet wins of a well-run finance and accounting function.

The obvious next question is how much of this a machine can take off your hands.

How much of variance analysis can you automate?

The calculation, the threshold flagging, and a first draft of commentary all automate well, because each follows a rule. Identifying the real business cause and deciding what to do about it does not, because that needs context no ledger contains.

Automated commentary has improved a lot, and it is genuinely useful as a starting point. It can tell you professional fees rose because three new invoices hit a cost center.

What it cannot tell you is that those invoices exist because a migration slipped, that the slip was a scoping error, and that the same thing will happen again next quarter unless someone changes the process.

That gap between description and explanation is the whole argument for keeping analysts in the loop, and it mirrors what we found looking at what stays human when you offshore to India.

In practice the work divides fairly cleanly.

The pattern in that right-hand column is consistent across back-office functions, and we mapped how teams progress through it in our agentic offshoring maturity model.

If you are deciding which processes to move first, we scored them by readiness in our guide to which business functions to offshore to India.

So the practical question is who sits in that analyst seat, especially when the finance team is three people.

Who does variance analysis on a lean finance team?

On a small team it usually falls to the controller, squeezed in after close, which is why commentary is often the first thing to slip. The fix most companies land on is one dedicated analyst who owns the pack rather than adding it to someone's existing job.

That single hire changes the output more than any software purchase, because commentary quality depends on someone having time to ask operations what actually happened.

India is a common place to add that capacity, partly on cost and partly because the accounting talent pool is deep in qualified chartered accountants and MBA-trained analysts.

Where in India you build matters, and we compared the main options in our guide to the best Indian cities for offshore finance operations.

For a worked account of how the first few hires actually go, read how US startups build finance operations teams in India.

A working variance and reporting pod tends to cover four roles:

• FP&A analyst: owns the variance pack, drafts commentary, and chases the causes with operations.
• Reporting analyst: maintains the models and dashboards so the numbers arrive the same way each month.
• General ledger accountant: closes the books cleanly, because bad data produces confident, wrong commentary.
• Finance lead: sets thresholds, reviews the narrative, and presents it upward.

One analyst plus an existing controller covers a surprising amount, and our breakdown of the cost of an offshore finance team in India shows how the pod scales from there.

Smaller teams often fold reporting into a broader offshore accounting function rather than staffing FP&A separately, which works well until the commentary starts slipping again.

The route you use to employ them decides how quickly they start. Setting up your own Indian entity first is the slow path.

Setup time: EOR, 1 to 5 days. Your own entity, 3 to 6 months. Upfront cost: EOR, $0. Your own entity, $15,000 to $25,000.

Wisemonk, Employer of Record in India guide, 2026

For a single analyst or a pod of four, that timing difference usually settles the decision on its own. Our comparison of EOR vs entity in India sets out where the crossover point sits as headcount grows.

If you are planning around a close calendar, our hiring timeline in India breaks down how long each stage realistically takes.

To model it properly, our breakdown of the cost of an Employer of Record in India separates the service fee from the statutory load.

And if the model itself is new to you, our explainer on what an Employer of Record actually does is the clearest starting point.

Recruiting the analyst is the other half of the problem, and it is worth knowing what that costs before you start.

Recruitment concierge: 10% of annual salary. Access the top 1% of India talent, screened through rigorous assessment. 90-day placement guarantee. No upfront cost, pay on join date.

Wisemonk pricing page, 2026

Budget the fully loaded figure rather than the salary, since employer contributions sit on top, and our guide to the cost of employment in India sets out what those add.

For the wider context on why finance work moved to India at all, see our overview of India outsourcing.

If this would be your first hire there, start with our step-by-step guide to building an offshore team in India.

And for the strategic case rather than the mechanics, read our guide to offshoring to India.

Once the analyst is in place, the work shifts to running the relationship well, which our notes on offshore team management cover.

The habits matter more than the reporting line, and we collected the ones that make a difference in our tips for working with offshore teams in India.

How can Wisemonk help you build an FP&A team in India?

Wisemonk is an India-native Employer of Record (EOR) that helps global companies hire, pay, and manage talent in India without setting up a local entity.

For a reporting function, that means you can put a named FP&A analyst on your monthly pack within weeks, on a compliant Indian employment contract, without registering a company in India first.

We support 300+ global clients and more than 2,000 employees across India, process $20M+ in annual payroll, and hold a 4.8/5 rating on G2. Pricing starts from $99 per employee per month.

Here is how we help:

• Recruitment: we source and screen chartered accountants and FP&A analysts against your own reporting requirements.
• Managed payroll: monthly payroll in rupees with PF, ESI, professional tax, and TDS filings handled for you.
• Contractor management: if you want a fractional analyst for a few close cycles first, we contract and pay them compliantly.
• Background checks: analysts see your unpublished results, so we verify credentials and history before they start.
• GCC setup: when reporting grows into a wider shared services function, we help you build it out.
• Entity setup: if the team reaches the scale where your own Indian entity makes sense, we support that move.

From our experience helping companies build finance teams in India, the analysts who produce the best commentary are the ones given direct access to operations leads rather than kept inside the finance function.