What is a rolling forecast, and should you use one?
A rolling forecast always looks the same distance ahead: each time a period closes, you drop it and add a new one at the far end, so a 12-month rolling forecast is always 12 months long. It replaces the annual budget's shrinking horizon with a constant one.
A traditional annual budget is set once and then gets shorter every month. By November you are managing a business with a two-month planning horizon, which is the point in the year when you can least afford one.
A rolling forecast fixes the horizon rather than the end date. When January closes you add the following January, so you are always looking twelve months out. Four, six and eighteen months are all used; twelve is the most common.
What it is good for#
It keeps the forecast honest. An annual budget becomes fiction some months in, and everybody knows it, but it is still the document being reported against. A rolling forecast absorbs what actually happened as it happens, so the number you are managing to reflects the business you actually have.
It also removes the annual planning cliff. Instead of one exhausting exercise every autumn that everybody games, planning becomes a monthly adjustment. And it makes decisions easier to time, because a hire in month eleven is always visible rather than falling off the end of the budget.
What it costs#
Effort, every month, forever. This is the honest objection and the reason most attempts are abandoned. If updating the forecast means a day of somebody rebuilding spreadsheet links, a rolling process will not survive contact with a busy quarter.
There is also a governance question. If the forecast moves every month, what are people accountable to? Most organisations answer this by keeping an annual target for accountability and running the rolling forecast alongside it for decisions. Blurring the two is how a rolling forecast quietly becomes a way of never missing a number.
Who benefits most#
Businesses where conditions change faster than a year: early-stage companies, seasonal businesses, anyone growing quickly, anyone in a volatile market. A stable business with predictable annual rhythms gets less from it, and should not adopt it out of fashion.
How to start#
Start with fewer drivers than you think you need. A rolling forecast built on fifteen assumptions that get genuinely reviewed each month beats one built on two hundred that get copied forward untouched. Then set a fixed monthly slot for the update, and compare last month's forecast against actuals before you change anything. The variance is where the learning is.