Revenue came in exactly on forecast.
Not close. Exactly.
The sales team is celebrating. The revenue forecast accuracy dashboard is green. Someone is probably preparing a PowerPoint slide about improved forecasting discipline.
There’s just one problem.
Operating profit missed forecast by $600,000.
And now the CFO wants to know how finance managed to forecast the top line perfectly while getting the bottom line so spectacularly wrong.
It’s a fair question.
Revenue forecast accuracy doesn’t guarantee profit forecast accuracy. A company can correctly predict total sales while getting product mix, contribution margins, labor requirements, and operating costs wrong.
Those errors can offset one another at the revenue level while accumulating in operating profit.
The solution isn’t necessarily a more complicated forecasting model.
It’s understanding which assumptions actually determine profitability and testing whether those assumptions still reflect how the business operates.
How Revenue Can Hit Forecast While Operating Profit Misses by $600,000
Consider a hypothetical company selling two products.
Product A generates a 50% contribution margin.
Product B generates a 20% contribution margin.
Contribution margin here means revenue less the variable costs attributable to those sales. Fixed operating expenses are deducted separately to calculate operating profit.
Finance forecasts $10 million in monthly revenue, with most sales coming from the higher-margin product.
Actual revenue reaches $10 million.
But customers purchase considerably more of Product B than expected.
Here’s what happens.
| Measure | Forecast | Actual |
|---|---|---|
| Product A revenue | $6,000,000 | $4,000,000 |
| Product B revenue | $4,000,000 | $6,000,000 |
| Total revenue | $10,000,000 | $10,000,000 |
| Product A contribution margin | 50% | 50% |
| Product B contribution margin | 20% | 20% |
| Total contribution | $3,800,000 | $3,200,000 |
| Fixed operating expenses | $2,000,000 | $2,000,000 |
| Operating profit | $1,800,000 | $1,200,000 |
Illustrative assumptions: Product contribution margins remain constant, fixed operating expenses are unchanged, and there are no other operating income or expense items.
Revenue forecast error: $0.
Operating profit forecast error: −$600,000.
The entire difference comes from product mix.
Finance expected $6 million in sales from a product generating 50 cents of contribution for every revenue dollar.
Instead, $2 million of those sales shifted to a product generating 20 cents.
The calculation is:
Product mix impact = Revenue shifted × Difference in contribution margin
$2,000,000 × (20% − 50%) = −$600,000
No revenue was lost.
No additional operating expenses were incurred.
The company simply earned less contribution from the revenue it generated.
The revenue forecast was accurate in aggregate.
The underlying commercial assumptions were not.
That’s an important distinction when the CFO asks whether the forecasting process is working.
Revenue Accuracy Can Hide Two Forecasting Errors at Once
Look at the product-level results.
Product A missed its revenue forecast by $2 million.
Product B exceeded its forecast by $2 million.
The errors cancel at the consolidated revenue level.
This is why aggregate forecast accuracy can be misleading.
Using the example:
Net revenue forecast error = Actual revenue − Forecast revenue
$10,000,000 − $10,000,000 = $0
But consider the absolute errors by product.
Product A: $2,000,000
Product B: $2,000,000
Total absolute product-level error: $4,000,000
The company has $4 million of absolute forecast error across its two product lines despite a zero consolidated revenue error.
One useful diagnostic is weighted absolute percentage error, calculated as:
WAPE = Sum of absolute forecast errors ÷ Sum of actual values
$4,000,000 ÷ $10,000,000 = 40%
That means the total absolute product-level revenue error equals 40% of actual consolidated revenue.
It doesn’t mean consolidated revenue was off by 40%. It wasn’t off at all.
It means the forecast was materially wrong about where revenue would come from.
That distinction is useful because different products can carry different margins, working capital requirements, production constraints, and service costs.
The Association for Financial Professionals’ financial modeling materials discuss the limitations of averaging signed forecast errors, where positive and negative differences can cancel, and the usefulness of absolute-error measures.
FP&A should evaluate forecasting accuracy at the level where errors change decisions.
For a business with materially different product economics, that may be the product family, customer segment, channel, or geographic market.
A consolidated revenue forecast can look excellent while the forecast assumptions underneath it are quietly having a terrible month.
Separate Revenue, Mix, Margin, and Operating Cost Errors
The next step is determining whether the profit miss came from sales activity, product mix, unit economics, or operating expenses.
These drivers should be analyzed separately.
Otherwise, finance risks explaining the same financial movement multiple times.
Start with a simple operating profit relationship:
Operating profit = Revenue × Weighted contribution margin − Fixed operating expenses
For the example company:
Forecast:
$10,000,000 × 38% − $2,000,000 = $1,800,000
Actual:
$10,000,000 × 32% − $2,000,000 = $1,200,000
The weighted contribution margin declined by six percentage points.
At $10 million of revenue, each percentage point represents $100,000 of contribution.
Six percentage points therefore explain the $600,000 profit miss.
The challenge is determining why the weighted margin changed.
Was it product mix?
Did selling prices change?
Did material or freight costs increase?
Were discounts higher than expected?
Did labor productivity deteriorate?
In this particular example, the individual product contribution margins were unchanged.
That isolates the problem to product mix.
In a real business, several drivers may change simultaneously.
Finance needs a reconciliation method that distinguishes those effects without double-counting them.
Use a Profit Forecast Bridge That Actually Reconciles
A practical diagnostic starts with forecast operating profit and reconciles it to actual operating profit.
For a simple business, the bridge can separate:
- Total revenue movement at the forecast contribution margin
- Sales mix movement at forecast product margins
- Changes in product-level contribution margins
- Changes in fixed operating expenses
For the original example, the bridge is straightforward.
| Operating profit bridge | Impact |
|---|---|
| Forecast operating profit | $1,800,000 |
| Total revenue variance | $0 |
| Product mix variance | ($600,000) |
| Product contribution margin variance | $0 |
| Fixed operating expense variance | $0 |
| Actual operating profit | $1,200,000 |
The bridge reconciles exactly.
A useful calculation sequence for a business with multiple products is:
Step 1: Total revenue effect
(Actual total revenue − Forecast total revenue) × Forecast weighted contribution margin
Step 2: Product mix effect
Sum of (Actual product revenue − Actual total revenue × Forecast product revenue share) × Forecast product contribution margin
Step 3: Product margin effect
Sum of Actual product revenue × (Actual product contribution margin − Forecast product contribution margin)
Step 4: Fixed operating expense effect
Forecast fixed operating expenses − Actual fixed operating expenses
Under the stated model, the four effects reconcile forecast operating profit to actual operating profit.
This sequence deliberately measures revenue volume first, then mix at actual total revenue, then changes in product economics.
That matters because changing the calculation order can change how interaction effects are allocated.
There isn’t one universally correct variance decomposition for every business model. Finance should select a consistent methodology, document it, and ensure the bridge reconciles.
For a broader discussion of reconciling forecast versions, see 6 Checks Every FP&A Team Should Make Before Presenting a Forecast Bridge.
That process explains changes between successive forecasts. The analysis here answers a different question: why actual profitability departed from the forecast that management relied upon.
Check Whether Your Margin Assumptions Still Reflect the Business
Product mix isn’t always the problem.
Sometimes finance forecasts the right revenue by product but uses the wrong contribution margins.
Suppose a company expects $8 million in revenue at a 40% contribution margin.
It forecasts $3.2 million in contribution.
Actual revenue is exactly $8 million.
But actual contribution margin is 36%.
Actual contribution becomes:
$8,000,000 × 36% = $2,880,000
The contribution shortfall is:
$2,880,000 − $3,200,000 = −$320,000
Again, the revenue forecast is perfect.
The economics are not.
Finance now needs to determine whether the four-percentage-point margin decline came from selling prices, discounts, variable input costs, labor efficiency, freight, or other product-level expenses.
Don’t automatically assume that every margin decline represents an operational failure.
For example, management may deliberately accept lower margins to enter a new market, retain an important customer, or improve production utilization.
The forecast still needs to reflect that decision.
If the sales organization agrees to a discount in September and FP&A continues using the old pricing assumptions in October, the resulting profit miss isn’t mysterious.
It’s an information problem.
The model wasn’t updated to reflect a commercial decision that had already been made.
Don’t Treat Every Operating Expense as a Percentage of Revenue
Another common forecasting weakness is using revenue growth to predict expenses that don’t move proportionately with revenue.
Consider labor costs.
A company forecasts $12 million in revenue and $3 million in labor expense.
Its model assumes labor will remain at 25% of revenue.
Actual revenue reaches $12 million.
But labor expense increases to $3.4 million.
The company misses its operating profit forecast by $400,000 before considering any other changes.
What happened?
Possibilities include overtime, staffing mix, wage increases, training costs, lower productivity, or additional capacity requirements.
A 25% labor-to-revenue assumption doesn’t explain any of them.
It simply assumes the relationship will remain constant.
That might be reasonable for certain variable compensation expenses.
It may be inappropriate for salaried employees, step-fixed staffing requirements, or operations with changing productivity.
A stronger labor forecast could use:
Required labor hours = Forecast production volume × Standard labor hours per unit
Labor expense = Required labor hours × Expected loaded hourly cost
Then separately model salaried staff, overtime premiums, vacancies, temporary labor, and other material components.
For example, suppose production requires 100,000 units and the standard is 0.5 labor hours per unit.
Expected labor hours:
100,000 × 0.5 = 50,000 hours
At a loaded hourly cost of $35:
50,000 × $35 = $1,750,000
If actual productivity deteriorates to 0.55 hours per unit, with volume and hourly cost unchanged:
100,000 × 0.55 × $35 = $1,925,000
The resulting $175,000 unfavorable labor variance is explained by productivity.
Management can now investigate the cause.
Perhaps production downtime increased.
Perhaps a new product requires additional handling.
Perhaps the standard was unrealistic.
Those explanations lead to different decisions.
“Labor was 10% over forecast” doesn’t tell anyone what to fix.
Decide Which Forecast Drivers Deserve Their Own Models
Driver-based forecasting can improve visibility, but complexity isn’t automatically an improvement.
FP&A Trends discusses this problem in Driver-Based Forecasting in FP&A: Common Pitfalls and Practical Solutions, including the risks of excessive detail and disconnected models.
A business doesn’t need a separate forecasting model for every expense account.
It needs enough detail to identify the assumptions that materially affect financial outcomes.
A practical test is to ask three questions about each proposed driver.
Does it materially affect operating profit?
A driver responsible for a $500,000 potential profit movement deserves more attention than one that changes the forecast by $2,000.
Materiality should reflect the company’s scale, earnings volatility, and decisions being made.
Can the relationship be explained and tested?
A driver should have a credible economic connection to the forecast outcome.
Historical correlation alone doesn’t establish causation.
For example, travel expenses and revenue may both increase during a growth period. That doesn’t necessarily mean revenue growth directly causes every travel expense.
Will better forecasting of this driver change a management decision?
If improved detail won’t affect pricing, staffing, purchasing, capacity, investment, or another meaningful decision, consider whether the additional maintenance is worthwhile.
A sophisticated model that takes eleven days to update and still cannot explain margin deterioration has achieved an unfortunate combination of complexity and inconvenience.
Test Forecast Accuracy at More Than One Level
Finance should stop relying on a single forecast accuracy percentage to judge whether the forecast is useful.
Different metrics answer different questions.
| Measure | What it reveals |
|---|---|
| Consolidated revenue error | Whether total revenue matched the forecast |
| Product-level absolute error | Whether sales were correctly distributed across products |
| Weighted contribution margin error | Whether forecast revenue produced the expected contribution |
| Operating expense error | Whether spending assumptions were reliable |
| Operating profit error | Whether the combined forecast correctly predicted profitability |
| Signed forecast bias | Whether errors consistently lean optimistic or pessimistic |
These measures should be evaluated across multiple periods.
One accurate month may be luck.
One inaccurate month may reflect an unusual event.
Repeated patterns are more informative.
Suppose the company overestimates operating profit in five consecutive months.
That doesn’t prove intentional optimism.
But it should prompt an investigation into whether the forecast systematically assumes stronger margins, better productivity, lower spending, or more favorable product mix than the business actually delivers.
Also, be careful with percentage-based accuracy measures when actual values approach zero or become negative.
Operating profit percentage errors can become misleading or undefined around break-even.
In those situations, absolute dollar errors, margin-point differences, and consistently defined scaled measures may provide a clearer picture.
The purpose is to identify forecasting weaknesses, not award the model a flattering percentage.
Forecast Accuracy Should Change How the Next Forecast Is Built
Once FP&A identifies the source of a profit miss, the next question is what to change.
If product mix caused the problem, finance may need a better product-level demand forecast.
If discounts caused it, sales and finance may need a more reliable process for incorporating pricing decisions.
If labor productivity caused it, the operating forecast may need updated production standards and capacity assumptions.
If fixed operating expenses caused it, the issue may be weak departmental inputs, unrecorded commitments, or unrealistic spending assumptions.
The response should match the cause.
Don’t rebuild the entire forecasting model because one material assumption failed.
And don’t add another approval layer simply because the CFO is unhappy with the result.
The useful corrective action is the one that improves the information used to make the next decision.
For each material forecast miss, FP&A should document the underlying driver, evidence of the error, responsible assumption owner, proposed correction, and expected effect on future forecasts.
Then test whether the correction actually improves subsequent results.
A new assumption isn’t better merely because it is newer.
It needs to produce a more reliable forecast or improve the decisions management can make.
The CFO Needs to Know Whether the Business Is More or Less Profitable Than Expected
Return to the original company.
Revenue was exactly on forecast at $10 million.
Operating profit missed by $600,000 because sales shifted toward the lower-margin product.
That tells management something commercially important.
The business generated the expected revenue but earned less from it.
The next conversation should focus on why customers bought a different mix of products, whether that shift is likely to continue, and what management can reasonably do about it.
Should sales incentives change?
Is the higher-margin product losing demand?
Are customers responding to pricing differences?
Does the company need to adjust production capacity or its commercial strategy?
Or is the lower-margin product strategically valuable for reasons the current operating profit measure doesn’t capture?
Finance shouldn’t assume that restoring the old mix is automatically the correct decision.
It should calculate the economics of the available choices.
That is what makes forecast accuracy analysis useful.
A revenue forecast can be correct for the wrong reasons.
A profit forecast can be wrong for reasons management urgently needs to understand.
The job of FP&A is to distinguish between the two.
Before celebrating another accurate revenue forecast, ask one more question:
Did we correctly forecast how much money the business would actually make from those sales?
If the answer is no, the green dashboard can wait.
The Schlott Company’s Forecasting & Planning services help finance teams connect operating assumptions to financial outcomes, challenge forecast drivers, and build forecasts that management can use to make commercial decisions.