For accounting firms

Budget vs actual: the report, and the question it does not answer

A budget vs actual report places what the business planned beside what actually happened, line by line, and shows the difference in dollars and as a percentage. It is the most common management report a firm produces for a client, it is built into QuickBooks Online and Xero, and it is the first thing most owners are shown when a month has gone wrong.

In the practices we work with, the report gets discussed every month, but the explanation does not always carry into the next decision. This page covers what the report is, how to read it, and the three things it does not tell a firm that a client’s decision depends on.

What the report shows

Every row is an account or a category. Every row has the same four columns.

Budget. What the plan said this line would be for the period, usually a month, sometimes a quarter or the year to date.

Actual. What the books say it was, once the period is closed and the transactions are categorised.

Variance. The difference between the two, in dollars. Conventions differ; the one to state at the top of the report is the one that makes favourable and unfavourable easy to read. The example below shows favourable variances as positive and unfavourable in brackets. For revenue and operating result, variance is actual less budget. For expenses, it is budget less actual.

Variance percent. The dollar variance divided by the budget. This is what makes a 4,000 dollar miss on a 200,000 dollar line look different from a 4,000 dollar miss on an 8,000 dollar line, which it is. Where the budget for a line is zero, the percentage is undefined; show the dollar variance and N/A.

One requirement before any of this means anything: budget and actuals have to be on the same accounting basis. A budget built on cash timing compared with accrual actuals produces variances that are artefacts of the mismatch rather than signals about the business.

A short example, one month, for a service business:

Budget versus actual for one month. Variances in brackets are unfavourable.
LineBudgetActualVariance%
Revenue240,000226,000(14,000)(5.8%)
Cost of delivery96,00091,0005,0005.2%
Payroll and contractors68,00079,500(11,500)(16.9%)
Rent and facilities12,00012,00000.0%
Marketing9,0004,2004,80053.3%
Other operating15,00016,100(1,100)(7.3%)
Net operating result40,00023,200(16,800)(42.0%)

Variances in brackets are unfavourable. The report is doing its job: it is immediately clear that revenue came in short, payroll ran well over, and marketing was underspent. The net result is less than sixty percent of plan.

Which variances are worth explaining

Not every line deserves a sentence. The useful discipline is a threshold, set per client, above which a variance gets a written explanation and below which it is noted and left alone. Ten percent and a dollar floor is a common starting point; the right numbers depend on the size of the line and how much the owner will act on it.

Above the threshold, the review asks two questions of each variance, and they are different questions.

What changed. Is this timing, volume, price or another driver? Timing means the activity landed in a different period from the one the plan assumed: a project was delivered and its revenue recognised in July rather than June. Volume means the business did more or less of something: two fewer projects closed. Price means it paid or charged a different rate: a supplier’s cost fell, or cover was bought at a contractor’s rate rather than an employee’s. A timing variance is expected to reverse, and that expectation still needs checking in the following period, because a shift in timing can carry a cash consequence even when the accrual result recovers.

Why it changed. Was this a deliberate decision, an external event, or an assumption that did not hold? The payroll overrun in the example is a rate variance in the first question and a decision in the second: a technician resigned in March and the owner brought on a contractor at a higher rate to cover the work. The marketing underspend is reduced activity in the first question and a decision in the second: the campaign was paused in April. Both were reasonable calls. The revenue shortfall may be an assumption that did not hold, if the plan expected a renewal that slipped.

The report flags where to investigate. Supporting detail explains what changed. A record kept at the time preserves why a decision was made and what would warrant revisiting it. It is the second question that decides what the owner should do next.

Where the standard report stops

A budget vs actual report answers one question well: where did the business depart from the plan this period. Three related questions decide what the client should do next, and the report is silent on all of them.

It shows the departure, not where the business now lands

Revenue is 14,000 short this month. The report says so. What it does not say is whether that shortfall is a one-month dip or the first month of a new run-rate, and what the year now looks like if it is the second.

That question needs the forecast beside the budget, not just the actuals. Budget-to-actual shows what has already happened. Forecast-to-budget shows where the accumulated variances lead. A firm running only the first sees each month’s miss and not its consequence. A firm running only the second has quietly redefined the target as whatever it now expects.

The two documents have different jobs and change on different schedules. The budget is held fixed so variance has a reference point. The forecast moves whenever the information changes. Budget vs forecast: what the difference is for covers why both are needed and what the third thing is that a client’s affordability question requires.

It shows the line, not the decision that moved it

Payroll is 11,500 over. The report shows the symptom. The cause is a decision made in March, when a technician resigned and the owner brought on a contractor at a higher rate to cover the work. That decision was judged against March’s cash position, which was comfortable.

By the time the June report shows the overrun, nobody is prompted to ask whether the March decision still holds, because the report has no record that a decision was made, what it was judged against, or that the position it was judged against has moved. The variance is explained, the explanation is delivered, and the same line runs over next month.

The fix is a decision record kept beside the budget: date, the commitment, what the plan said, what the forecast said at the time, and the reason for going ahead. With that record, a variance can be traced to a decision rather than only to a line, and the decisions that need revisiting when cash moves are a list rather than a memory. Why the budget you built stops getting used covers the four failure modes that follow when the record does not exist.

It only means something if the plan stayed still

A variance is the difference between what was intended and what happened. It is only legible if the intention did not move. When a budget is revised month by month to track the forecast, the two numbers converge, the variance shrinks, and the business has lost the ability to be surprised.

That is the trap in the standard advice to keep the budget current. Review budget-to-actual monthly. Update the forecast whenever material information changes. Revise the approved budget only when the business deliberately changes its plan, as a dated version with the prior one kept. How often should you update the budget covers why this discipline is safe inside a large planning system and why it usually is not when imported into a spreadsheet.

The assumption behind each material line belongs in the same place. Revenue of 240,000 rested on a renewal, a seasonal pattern or a hire ramping. When the month comes in at 226,000, the assumption is what says which part of the reasoning failed. Without it, the variance is a number with no explanation attached, and next year’s budget carries the same assumption forward unchanged. What your forecast learns from being wrong covers how the grading works.

What this means for the firm

The report is not the problem. It is a good report and it should be produced. What the practices we work with describe is a report that is produced, discussed, and then has nothing to do with the decision the owner makes the following Tuesday, because the plan lives in one file, the forecast in another, and the decision is made against the bank balance.

A budget vs actual review that does its job has three things beside it: the current forecast, so the departure has a consequence; the decision record, so the departure has a cause; and the assumptions, so the departure has a lesson. Each of those is a small amount of work for one client. Across a portfolio, they are the work that decides whether the budget is an instrument or a record of January’s intentions. What a small business budget is actually for opens the series with what the document is doing at this size.

The gap is in the structure, not the firm. A partner who remembers why every client’s payroll line moved is doing the work a system should hold.

Where AgentLink fits

AgentLink holds the approved budget as loaded, without rewriting it, and takes a new version only when the firm deliberately supplies one. It keeps the forecast current from the bank and the ledger daily. When an owner asks whether something is affordable, the answer is given against both the plan and the current forecast, and the decision is recorded as a by-product of answering, with the position it was judged against. When that position moves, the decisions judged against the old picture are surfaced for review. At year end the firm has the graded plan and the reasons the business moved away from it.

Whether to revise the plan, and what to advise the client about a variance, remains the firm’s judgement.

The budget moment describes how this fits the engagement.

Frequently asked questions

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

They are the same comparison at different depths. A budget vs actual report shows the differences. Variance analysis explains them: what changed, whether timing, volume or price, and why it changed, whether a decision, an external event or an assumption that failed. The report is the input; the analysis is the work.

How do I run a budget vs actual report in QuickBooks Online or Xero?

Both platforms let a budget be entered by account and month and compared with actuals in a standard report. In QuickBooks Online it is under Reports, Budgets vs Actuals. In Xero it is the Budget Variance report. Both require the budget to have been entered first. Notes and commentary can be attached in either. What neither maintains is a linked history of the decisions made against the plan, the position each was judged against, and a prompt to revisit them when that position moves.

Is there a budget vs actual Excel template I should use?

Any template with four columns per line will do: budget, actual, variance and percent. The layout is not where the value is. The value is in the threshold, the explanation above it, and the assumption and decision notes kept alongside, none of which a template supplies.

How often should the report be produced?

Monthly, once the period is closed and the transactions are categorised. Producing it before the books are settled produces variances that are reconciliation noise rather than business signal.

What is a good variance threshold?

Ten percent and a dollar floor is a common starting point. The right threshold is the one above which the owner would change a decision. A two percent variance on payroll may matter more than a forty percent variance on office supplies.

Should the budget be updated when actuals come in different?

No. The budget is the fixed reference point. Update the forecast to reflect what is now expected. Revise the budget only when the business deliberately changes its plan, and keep the previous version so variance can still be reported against the plan that was in force at the time.

What should the written commentary say?

For each variance above the threshold: what the plan assumed, what changed and whether it is timing, volume or price, and why it changed, whether a decision, an external event or an assumption that did not hold. One or two sentences each. Commentary the client will read is short and names the cause.

AgentLink builds controller-layer software sold through CPA and fractional CFO firms. The practitioner observations in the linked guides come from design-partner conversations and are used with permission where attributed. Corrections to nilanjan@agentlink.finance.