A project-based firm can be profitable on paper and still hit a cash crunch that forces ugly decisions — delayed payroll, emergency credit draws, damaged vendor relationships. The cause is almost always timing, not economics. Milestone billing lands in lumps while payroll goes out every week with mechanical reliability. The gap between those two rhythms is where liquidity surprises live, and a monthly P&L will never show it in time.
Why 13 weeks
The 13-week horizon — roughly one quarter — is a practitioner convention because it matches the useful prediction window. Less than a quarter, and you're just managing the payables schedule; more than a quarter, and you're guessing at work that hasn't been sold and clients whose payment behavior you can't predict. Thirteen weeks is far enough to see a crunch forming and close enough that receipts are grounded in real milestones, real receivables, and real billing schedules. It's updated weekly, and each week the completed week drops off and a new week thirteen is added, so the horizon never goes stale.
The weekly grain matters as much as the horizon: a firm can look fine monthly while going negative in week 3. Project firms live on weekly payroll cycles, so the forecast should too.
What goes in
The forecast has two halves: receipts and disbursements, dated by the week cash actually moves.
Receipts are broken down by client and milestone, not aggregated into "collections." For each expected payment, record the client, the project or invoice, the amount, and — critically — the week you expect the cash to arrive. Ground each date in the receivables aging and the billing schedule: invoices already sent get a payment week based on the client's actual payment history, not the printed terms; milestone receipts go in the week the billing will realistically be approved and paid. Put a milestone that "might slip" in the conservative week — optimism is the single biggest source of forecast error.
Disbursements are the weekly outflows that don't care about your billing cycle: payroll and benefits timed to your actual pay dates (always the largest line), subcontractor and vendor payments, rent and facilities, debt service and equipment payments, and operating expenses. The discipline here is placing lumpy payments in their exact week — the quarterly insurance premium, the annual software renewal, the estimated tax payment — because a forecast that spreads them evenly is lying to you in the weeks that matter most.
Each week nets to an opening and closing cash balance, with a summary row tracking cumulative net flow and a minimum-threshold line — many firms use four to eight weeks of disbursements — so dips below it show at a glance. The closing balances across the 13 weeks are the whole point: one line that tells you where cash is heading.
The weekly discipline
The forecast is not a document; it's a rhythm. Thirty minutes, same time each week, with the people who control billing and spending. First, replace last week's forecast columns with actuals and note the variance drivers — the slipped receipt, the early payment, the payroll run that was higher than expected. Second, re-forecast receipts: walk the aging and the milestone schedule, confirm the week each payment lands, and move anything that slipped. Third, re-forecast disbursements: confirm payroll, upcoming vendor payments, and any new one-off outflows. Fourth, read the cash line across all 13 weeks and act on the weak weeks now — accelerate a milestone billing, pull in a client payment, defer a discretionary spend, or arrange a credit line draw before the need is urgent. Fifth, roll the horizon forward a week.
Common mistakes
The first is forecasting invoice dates instead of cash dates. The forecast only cares when cash lands: a milestone invoiced this week and collected in five weeks belongs in week five, not week one. Putting it in week one manufactures a false sense of safety.
The second is hiding the minimum threshold. A forecast with no stated floor always looks fine; the same forecast against a four-week-disbursement minimum shows exactly which weeks need action. The threshold isn't a covenant — it's a decision trigger.
The third is treating it as a finance exercise. If only the bookkeeper sees the forecast, nothing changes. The value comes when the operations lead sees a tight week seven and moves a milestone billing forward, or when the owner sees a strong position and brings a hire forward. Behavior changes when the decision-makers are in the room.
The fourth is letting it go stale. A forecast updated "when we get to it" is a document; a forecast updated weekly is a system. The week it goes stale is usually the week it would have warned you.
How it changes decisions
A working 13-week forecast changes the questions leadership asks. Instead of "can we make payroll?" — a question asked in panic — the firm asks "week seven is tight; which milestone can we accelerate?" Instead of discovering a cash dip when the balance goes low, the firm sees it six weeks out, when options are cheap: pull a payment in, defer a spend, schedule a credit draw calmly. Hiring decisions get timed to the forecast rather than to optimism. Large purchases get placed in strong weeks. Lenders get a document that demonstrates control instead of a story told after the fact.
None of this requires sophisticated finance. It requires one spreadsheet, honest dates, and a weekly thirty-minute meeting that never gets skipped. The firms that run into cash trouble are rarely the ones with bad economics — they're the ones that managed cash by checking the balance. The 13-week forecast replaces the balance check with foresight, and foresight is the difference between reacting to a crunch and planning around it.
See the free 13-week cash flow forecast template — the walkthrough covers the weekly receipts and disbursements structure, the opening-to-closing cash mechanics, and the variance discipline described above.