Margin and cash are different problems
A commercial contractor can hold a healthy backlog, report reasonable margins, and still be unable to meet payroll, because construction payment timing is brutal. You pay labor weekly, pay suppliers on net terms, bill monthly, get paid somewhere between thirty and ninety days later, and a portion is withheld as retainage until long after the work is complete.
The gap between spending and collecting is funded by the contractor. Growth widens the gap, which is why contractors most often fail while expanding. Forecasting this gap is not an accounting nicety. It determines how much work the company can safely carry.
Build the forecast from project schedules
A useful forecast starts at the project level. For each active job: planned cost by month from the schedule and budget, expected billing by month, the historical payment lag for that particular general contractor, and the retainage percentage with its expected release timing. Aggregate across jobs and the company picture emerges.
Payment lag by client is the input most often assumed rather than measured. Contractors usually know which general contractors pay slowly, but rarely have the number. Calculating the actual average days from submission to payment for each client, from the records you already have, sharpens the forecast more than any other single improvement.
Include the items that are easy to forget
Several predictable cash events routinely go missing from forecasts. Retainage accumulating across multiple jobs and releasing unpredictably. Unapproved change order costs being carried while approval is pending. Material deposits and progress payments on long-lead equipment. Mobilization costs on new jobs that precede any billing. Insurance and bonding premiums that arrive in lumps.
Each of these is knowable in advance. Pulling them from the change order log, the procurement record, and the project start schedule into the forecast makes the difference between a projection that holds and one that is wrong in the specific month it mattered.
Update weekly, automatically
A forecast built by hand in a spreadsheet is accurate the day it is built. Construction changes faster than that. A job starts late, a change order gets approved, a payment arrives early, a supplier demands different terms. By the time a monthly forecast is reviewed, several assumptions behind it are no longer true.
Pull the inputs automatically from the systems that hold them and refresh weekly. The forecast does not need to be precise to be useful. It needs to be current enough that when the projection shows a tight month eight weeks out, there is still time to draw on a line of credit, accelerate a billing, or negotiate terms. That lead time is the entire value of forecasting.
Know your safe backlog
Growth consumes cash before it produces any. Each new project requires mobilization, material purchases, and payroll well before the first payment arrives, and several starting close together can exceed what the company can fund even when every job is profitable. This is the mechanism behind most contractor failures during expansion.
Use the forecast to establish how much new work the company can start in a given period given its cash position and credit availability. Then treat that limit as a real constraint in the go or no-go decision rather than an accounting opinion. Declining a good project because the timing does not fit is a difficult call, and it is considerably easier than the alternative.