Rolling Cash Flow Forecast

How a rolling forecast works

A traditional annual budget shrinks as the year passes — by Q4 you have only three months of forward visibility, exactly when you need the most. A rolling forecast solves this by adding a new period at the end every time one falls off the front.

The mechanics are simple: set a horizon (commonly 13 weeks for cash, 12 months for the broader plan), set a refresh cadence (Monday morning is typical for cash), and on each refresh drop the period that just ended, add a new period at the far edge, and update assumptions for everything in between using the latest actuals.

Why operators prefer rolling forecasts

Rolling forecasts force a discipline of continuous re-planning rather than annual heroics. They also surface variance early: if week 3's actual receipts are 20% below forecast, the next refresh propagates that into weeks 4-13, which often changes hiring, marketing, or financing decisions before the cash crisis hits.

The trade-off is workload. A weekly cash forecast refresh takes 30-90 minutes once data sources are wired in. Most owner-operators find that's a small price for the visibility — and treasury teams at PE-backed companies have run weekly 13-week forecasts for decades.

Maintenance discipline

A rolling forecast only earns its keep if the update cadence is religious. The standard for a 13-week model is a Monday or Friday refresh: actuals replace the prior week's plan, the horizon rolls forward by one week, and any new contracts, invoices, or commitments get added. Treat the update as a 30-minute meeting on the calendar, not a 'when we get to it' task — the moment you skip a week, confidence in the numbers collapses.

Forecast variance is the most useful artifact the process produces. Track plan-vs-actual by line item each week and look for systematic bias: collections always 5 days late, payroll always $2k higher than budgeted, a particular customer that consistently slips. Within two quarters this becomes the foundation of a much more accurate forecast because you stop forecasting the world you wish you lived in and start forecasting the one you actually run.

Tools matter less than cadence. A clean spreadsheet updated weekly outperforms a slick FP&A platform updated quarterly. Many CFOs run a 13-week direct-method model in Google Sheets or Excel and only graduate to dedicated software (Causal, Pry, Jirav, or a purpose-built planner) once headcount or transaction volume makes manual entry untenable.