Home / Learn

How do you explain a budget versus actual variance?

Check that budget and actuals cover the same measure, period and scope, then calculate the difference and follow it into the records. Separate changes in activity from changes in effort or cost per unit. Use the evidence to explain what happened and identify who will respond.

A team used more operating hours than planned. It may have completed more jobs, handled harder work or changed its process. The recorded total leaves all of those explanations open.

The review needs to establish which one fits. Follow the movement into jobs and transactions, then ask the person responsible for the process to interpret it. A useful explanation connects that evidence to a decision about what happens next.

Fix the basis of comparison

A useful variance compares the same measure, period and business scope against a named plan. Identify the original budget separately from any later forecast. Each establishes a different expectation against which to assess actual performance.

Check the entity, account and unit. Financial comparisons also need aligned currency and adjustment rules. A reorganization may change where activity is recorded without changing the activity itself.

Review timing before treating a difference as an operating problem. A late entry can affect an earlier close; planned work may have moved into the next month. Retained history helps distinguish those cases.

Keep the originally issued version identifiable when a comparison is revised. A manager may need the corrected view for today's decision and the earlier view to explain what was understood at the time.

Use a consistent variance calculation

Choose one sign convention and show what an increase means for the measure. Under actual minus plan, a positive variance means the actual exceeded the plan. It does not automatically mean performance improved.

For a positive, nonzero plan, divide the difference by the plan and multiply by 100 to express the percentage. With 108 actual hours and 100 planned hours, the difference is eight hours, or 8% above plan.

When the plan is zero, that percentage is undefined. Show the absolute movement and explain the new activity. A negative baseline also requires explanation because its percentage can be difficult to interpret.

Read the sign alongside the business circumstances. Lower expenditure might result from an improvement, but it could also reflect work postponed or a vacancy. The number needs that explanation before it supports a response.

Separate activity from performance

Calculate how much resource the actual workload should have required at the planned standard. That gives the manager a comparison which allows for a change in activity. The difference between this figure and actual resource use needs further explanation.

Below, the team planned 100 jobs and completed 110. The additional jobs explain some of its extra hours; the remaining hours require a review of the work.

Financial analysis makes related distinctions between volume and unit-cost effects. The SQA formula sheet separates usage and efficiency from other cost effects. Use the calculation suited to the question and check that its components add back to the overall variance.

A change in work mix needs further analysis. Ten complex jobs may require different effort from ten routine jobs. Holding the mix constant in a simple calculation makes the arithmetic clear, but a real review must examine whether that assumption fits.

Investigate the cause behind the movement

A variance calculation identifies a difference; the records and operating context explain it. Follow a material movement into the relevant jobs, transactions or assignments. Check whether the difference comes from activity, timing, classification or another recorded change.

A concentration in one department is a place to investigate. It does not establish the cause. The department may have taken on work previously recorded elsewhere, or its source may have received a late update.

Use the process owner's knowledge alongside the data. That person can explain circumstances the records capture only indirectly, such as a changed work sequence. Keep the observed movement separate from an explanation that still needs evidence.

Reconcile the drivers to the total. If the explanation accounts for only part of the variance, show the unresolved remainder. This prevents a plausible story from being mistaken for a complete reconciliation.

Connect the explanation to a decision

The report should make clear who will respond and what happens next. For each material variance, include the measure, period, plan version, actual, difference and explanation. Add the responsible owner and follow-up date where action is needed.

Choose review thresholds around the decision. A large percentage on a small base may matter less than a smaller movement in a critical process. Recurring differences may deserve attention even when each individual period falls below a threshold.

Commentary should explain whether the movement is temporary or likely to continue, with the evidence for that judgment. The proposed response may be to correct a record, change a process or revise an assumption in the next forecast.

At the next review, revisit the action. Record whether it happened and what changed afterwards. Otherwise, the organization can produce the same explanation every month without learning whether the response helped.

Make the review repeatable

Keep the source mappings, plan versions and refresh schedules used for the accepted comparison. Automating that preparation gives the next review a consistent starting point and leaves people time to investigate differences.

Check the inputs before interpreting the result. A successful refresh can still omit a late source. Show the missing period or stale record to both the operator and the person reading the report.

Databasin brings connected data and shared definitions into reports and AI questions, with implementation, rollout and ongoing support. The work can include the history needed for these comparisons. Finance can inspect how a variance was calculated, and the process owner can assess its explanation and choose a response.

Example: explaining additional operating hours

This illustrative month holds the mix of jobs constant. Each job has a standard allowance of two hours.

CalculationHoursInterpretation
Original plan: 100 jobs × 2 hours200The effort planned for the month.
Actual workload at standard: 110 jobs × 2 hours220The allowance for the jobs completed.
Recorded hours for those 110 jobs231The effort actually used.
Workload effect: 220 − 200+20Ten additional jobs account for these hours.
Remaining difference: 231 − 220+11The process owner needs to investigate these hours.
Total difference: 231 − 200+31The two components reconcile: 20 + 11 = 31.

The 31-hour increase is 15.5% above the original plan. The calculation separates its components; the manager still needs evidence to explain the remaining 11 hours.

Before the next variance review

  1. Align the comparisonName the plan, period, scope and unit. Establish how reorganizations and later corrections affect the figures readers will compare.
  2. Check the inputsVerify the expected period is available from each required source. Identify late or missing records before interpreting their effect as business movement.
  3. Reconcile the driversCalculate the activity component and investigate the remainder. Examine work mix and retain the portion that still needs an evidence-based explanation.
  4. Assign the responseState the decision, owner and follow-up for a material finding. Return to that action at the next review to see what changed.

Sources

Scottish Qualifications Authority, Formulae Sheet for Variance Analysis, 2019, version 1.4 · sqa.org.uk/sqa/files_ccc/AH_Accounting_FormulaeSheetVarianceAnalysis.pdf

Discuss your recurring variance review

We can talk through how your team explains differences and where preparation slows the review. A demo can explore the source data, historical definitions and reporting work that would make those discussions more useful.