You can be profitable on every single job, growing nicely, and still not make payroll. Profit gets recognised when the work is done. Cash turns up when the customer pays. Somebody has to fund the gap between those two moments, and if nobody has decided who, it turns into a crisis.
Growth makes it worse, not better. Every new job eats cash first. Materials, labour, subcontractors, all going out before anything comes in. A fast-growing business is constantly funding a bigger work-in-progress balance, which is why your worst cash months so often come right after your best sales months.
The thirteen-week rolling cash forecast is the standard tool for handling this. It isn’t a monthly cash flow statement. It’s a week-by-week projection of real receipts and real payments. This customer paying on that date. These bills falling due. Payroll dates, tax deposits, loan payments.
Thirteen weeks is the usual horizon because it’s long enough to give you room to move and short enough to still be roughly right. Push past a quarter and the timing of receipts turns into guesswork. Come inside a month and a problem you’ve just spotted may already be unavoidable.
The rolling update is what makes it useful. Each week you add a new week on the end and reconcile the week just gone against what really happened. That reconciliation is where the forecast gets better, because you learn which customers pay to terms and which ones don’t, and the projection sharpens up.
What you’re buying is time to decide. A squeeze you can see eight weeks out is manageable. Chase the collections, negotiate a payment date, draw on a facility because you chose to. That exact same squeeze spotted eight days out gets managed by whoever will lend you money fastest, on whatever terms they feel like.