Skip to content
Log in
← Glossary

rolling forecast

Also called: rolling 12-month forecast

A forecast updated every month or quarter that always looks the same distance ahead, instead of a fixed annual budget that goes stale.

A rolling forecast is updated on a fixed rhythm, usually every month or quarter, and always looks the same distance ahead. When a month closes, its forecast row is replaced with actuals, the owners of each driver say whether the coming months still hold, and one new month is added at the end. A twelve-month rolling forecast always shows the next twelve months, never a year-old guess.

It answers the main weakness of an annual budget: by spring the budget describes a world that no longer exists, yet decisions are still judged against it. The rolling forecast is the company's current best estimate, while the budget stays as the commitment results are compared with.

It works best on top of a driver-based model. Updating drivers (learners, price, payout, exchange rate) takes minutes and shows which assumption moved; re-typing totals takes days and explains nothing. Keep each update light: change only the drivers with new evidence, and note why, so the history of the forecast shows how the company's view changed.

Example

Klasna's forecast covers September 2026 to August 2027. September closes with 5,890 lessons and 996,149.75 UAH of contribution, against 5,612 lessons and 1,050,847 in the forecast.

The rolling update: the September row now holds actuals; the owners check the next months (the English price in Ukraine came in at 585 instead of 615 because of a promo: does it continue in October?); and September 2027 is added at the end. The forecast again shows twelve months ahead, October 2026 to September 2027, and the board sees the year's outlook move with the evidence instead of waiting for next autumn's budget.

Common mistakes

  • Re-forecasting totals instead of drivers. Without drivers an update can't say what changed.
  • Letting the horizon shrink. A forecast that runs only to December is a budget update, not a rolling forecast.
  • Replacing the budget with it silently. Keep both: the forecast is the best estimate, the budget is the commitment.
  • Changing every number every month. Update only drivers with new evidence and write down why.
  • Updating without owners. The people who run each driver confirm its next months.

Learn it in the course

  • Driver-based forecasting · Driver-based forecasting: revenue and profit from learners, lessons, price and payout instead of last year plus ten percent.