Budget vs actuals is the most routine deliverable in a fractional CFO engagement. It is also the one most commonly done superficially. A variance report that says "revenue was $50K below budget due to lower sales volume" is not analysis — it is a description. The analysis is why volume was lower, whether the shortfall is recoverable, and what the revised outlook means for cash and operations.
This article sets out a practical framework for variance analysis that goes beyond the numbers — covering how to structure the analysis, how to attribute variances to their root causes, and how to communicate findings to clients who are not financial experts.
Every material variance has three components that need to be addressed:
A variance explanation that addresses all three is genuinely useful to a board or management team. An explanation that only addresses the first is a description. Many "variance analyses" delivered by fractional CFOs are descriptions.
For revenue variances, the most useful attribution framework separates the total variance into three drivers: price, volume, and mix. This matters because the management response is different for each driver, and conflating them produces the wrong diagnosis.
The portion of the revenue variance attributable to selling at different prices than budgeted. Calculate as: (Actual price – Budget price) × Actual volume. If you sold 100 units at $95 instead of the budgeted $100, the price variance is ($95 – $100) × 100 = –$500. A negative price variance signals either pricing pressure from the market, customer-level discounting, or product mix shifting to lower-price tiers.
The portion attributable to selling more or fewer units than budgeted. Calculate as: (Actual volume – Budget volume) × Budget price. If you sold 90 units instead of 100 at the budget price of $100, the volume variance is (90 – 100) × $100 = –$1,000. A negative volume variance signals a sales execution problem, a market demand problem, or a capacity constraint — diagnosis requires looking at pipeline, conversion rates, and sales cycle data.
The portion attributable to the composition of sales being different from what was budgeted. Mix variance matters most when you sell multiple products or services with different margins. If you sold more of the lower-margin product than budgeted and less of the higher-margin product, revenue might be on plan but gross profit is below plan — the mix was unfavorable. Calculating mix variance requires margin data by product line, not just revenue totals.
The single most useful variance question: "If we had hit our budget volume at budget price, what would revenue have been?" This isolates the combined price and mix effect from the volume effect and tells you whether you have a pricing problem, a volume problem, or both.
For expense variances, the useful attribution splits between rate variances and efficiency variances — particularly for labour-heavy businesses.
A rate variance measures whether you paid more or less per unit of input than budgeted. If contractors cost $95/hour vs the budgeted $85/hour, the rate variance is ($95 – $85) × actual hours. An efficiency variance measures whether you used more or fewer units of input than expected for the output delivered. If a project took 120 hours instead of the budgeted 100, the efficiency variance is (120 – 100) × budget rate.
For non-labour expenses, the simpler attribution is between recurring and one-time items, and between controllable and non-controllable costs. A software subscription renewal at a higher rate is somewhat controllable; a property tax increase is not.
Gross margin compression is the variance that boards and investors watch most closely, because it tells them whether the fundamental economics of the business are improving or deteriorating. A 2-point margin decline that is one-time (due to a project cost overrun) requires a different response than a 2-point decline that is structural (due to labour inflation or product cost increases).
When gross margin is below budget, decompose it as follows:
Most fractional CFO clients are operators, not finance professionals. A variance analysis that leads with a table of numbers and follows with three paragraphs of accounting language will not land. The format that works best for non-financial audiences:
Variance analysis is only as reliable as the underlying data. If revenue recognition hasn't been properly applied, the revenue variance might be partly explained by a deferred revenue recognition difference, not actual commercial performance. If accruals haven't been run, the cost variances include timing noise that distorts the picture. If intercompany charges haven't been eliminated, the consolidated margin variance reflects internal transfers, not external economics.
A fractional CFO building variance analysis from accrual-adjusted, close-processed data can spend their full analytical time on the analysis. One building from raw QuickBooks data spends part of that time reconciling the data before the analysis can start.
Fynease Intelligence turns clean, close-processed financials into CFO-grade reporting — variance analysis, KPI dashboards, quality of earnings, forecasting, and board packs. Per client, for fractional CFOs managing QuickBooks Online companies.
See Fynease Intelligence Join waitlist