Revenue Forecast Variance

A firm forecasts $1,000,000 in monthly revenue and records $900,000 on the same accounting basis, so the variance is negative $100,000, or 10% below forecast, and the useful question is whether orders, prices, delivery timing or the model caused the difference. Confirm comparable definitions, because a forecast of invoiced sales cannot fairly be compared with recognised accounting revenue or cash received, and match period, currency, tax treatment and entity scope.

Calculate both amount and percentage: an illustrative variance is actual minus forecast revenue, which gives negative $100,000, and dividing by the forecast gives negative 10%, while a zero forecast needs a different presentation. Preserve the original forecast version, because if the team revises its forecast each week and later compares actuals with the final revision only, it may hide how early decisions were made, so report the version used for each decision.

CFO Perspective describes variance analysis as a way to compare results with forecasts and improve the next forecast, noting that model errors, missing data and human judgment can all cause differences. CFA Institute's company-forecasting guidance emphasises selecting drivers that fit the business model and testing assumptions, so a variance investigation should revisit those drivers rather than merely forcing future totals upward.

Break the gap into drivers such as units sold, realised price, discounts, product mix, customer segment and delivery timing, since a single total can hide offsetting positive and negative effects. Distinguish timing from lost demand, because a signed deal delivered next month may move revenue rather than remove it, depending on accounting rules, and the cash plan may be affected on a different date again.

Compare by product or channel, as a $100,000 shortfall in one line offset by a $90,000 surplus elsewhere produces a small net gap but may reveal a strategic change. Check data integrity, since a missing transaction feed, late journal entry or exchange-rate change can create an apparent operational variance, so reconcile actuals before attributing the result to sales.

Use a price-volume bridge with care: volume at planned price and price at actual volume can isolate two effects, but product mix and interactions need a documented allocation convention, and the bridge is not uniquely determined. Set materiality thresholds, because a small percentage on a huge base may matter while a large percentage on a tiny line may not, and review business consequence as well as the ratio.

Look for persistent bias, since forecasts that beat actuals every month may reflect overly optimistic deal dates or a weak churn assumption, while a single miss may be normal uncertainty. Record explanations with evidence that connects to orders, customer decisions or product performance, not just "market conditions", and avoid punishing honest forecasting because teams that fear a negative variance may deliberately sandbag.

Connect the finding to action, since a volume shortfall may change inventory orders, a pricing gap may prompt discount review and a timing shift may affect cash plans, and not every variance needs a new policy. Report uncertainty, so the next forecast shows the revised assumption and its risk and does not claim that one explanation guarantees a rebound.

For an owner, revenue forecast variance is a learning loop that tells how far reality moved from a plan, what drove the movement and whether decisions should change.