The 13-Week Cash Flow Forecast: A Treasury Discipline for Service Businesses

Walk into the finance office of almost any private-equity-backed company on a Monday morning and you'll find the same artifact on someone's screen: a single spreadsheet, thirteen columns wide, with rows for opening cash, weekly receipts, weekly disbursements, and ending cash. It's called the 13-week cash flow forecast, and it's been the dominant treasury discipline in mid-market finance for forty years.

The discipline is almost completely absent from the world of service businesses, freelancers, and agencies — even though their cash flow is far more volatile and their margin for error is far smaller. This whitepaper is a working blueprint for adopting it.

What the 13-week forecast actually is

The 13-week forecast is a rolling weekly view of every dollar entering and leaving the business over the next quarter. It is not a budget. It is not a P&L. It tracks cash — money landing in and leaving the bank account on a specific date — and nothing else.

A canonical layout has four sections, top to bottom:

1. Opening cash for the week — what you have in the operating account on Monday morning.
2. Receipts — every expected inflow: client payments, refunds, loan draws, owner contributions.
3. Disbursements — every expected outflow: payroll, contractors, rent, software, taxes, owner draws, debt service.
4. Ending cash = opening + receipts − disbursements. This becomes Monday's opening cash for the next column.

Every week, you drop the column that just ended and add a new column thirteen weeks out. The forecast always looks one quarter ahead.

Why thirteen weeks specifically

The number isn't arbitrary. Thirteen weeks is one calendar quarter — long enough to see the full receivables cycle of a service business (most invoices are net-30 to net-60, with some stragglers at net-90), and long enough to expose seasonal cash troughs that shorter horizons miss.

It is also short enough that the forecast remains specific. Most operators can describe what they expect to happen in the next ninety days with reasonable confidence: which clients owe what, which projects are kicking off, which expenses will hit. Past the 13-week horizon you're guessing about prospects, not invoicing receivables — and that's a different mental model. Keep them separate.

Inputs: where every line comes from

The data discipline is what separates a forecast you can trust from a fantasy. Every cell in the 13-week grid should trace back to a specific source document:

• Receipts from open AR — every unpaid invoice, dropped onto its expected payment date (terms + your client's actual payment behavior, not their stated terms).
• Receipts from work-in-progress — invoices you haven't sent yet but will send in the forecast window, dropped onto their expected payment date.
• Receipts from new sales — only the deals you'd commit to in front of your board. Use a probability tag and consider tracking a "low confidence" version separately.
• Recurring disbursements — payroll, rent, software, debt service, retainer-based contractors. These are the easy ones; build them once, reuse weekly.
• Variable disbursements — taxes (quarterly estimated payments hit specific dates), insurance renewals, equipment, travel, owner distributions.

A useful rule: if you can't name the customer, vendor, or invoice number that drives a line, the line doesn't belong on the forecast yet.

The Monday-morning refresh ritual

The forecast is only useful if it is refreshed weekly — same day, same time, same person. Most operators we've talked to spend 30 to 60 minutes on it. The ritual has four parts:

1. Reconcile last week. What did the bank actually do? Which receipts landed, which slipped, which disbursements surprised you?
2. Roll the forecast. Drop the just-ended column. Add a new column thirteen weeks out. Pull recurring lines forward.
3. Update the moving parts. Re-date receipts that slipped. Add new invoices issued. Reflect any new contracts or hires.
4. Identify the trough. Where is the lowest ending cash balance in the next thirteen weeks? Is it acceptable? If not, what action do you take this week?

That last step is the entire point. The forecast exists to drive a decision. If you reach the end of the refresh and haven't named at least one action — chase an invoice, defer a hire, draw on a line of credit, accelerate a sale — you didn't really refresh; you copied numbers.

Actuals vs. forecast: the variance column nobody builds

The single highest-leverage addition to a 13-week forecast is a variance column for the most-recently-completed week. Three numbers: forecasted ending cash, actual ending cash, difference. Then a one-line written explanation of the gap.

Over a quarter you'll discover patterns: client X always pays five days after they say they will; software renewals always show up two weeks later than the calendar invite; payroll taxes are always slightly higher than the gross figure suggests. Within ninety days the variance column is doing more for your forecast accuracy than any clever modeling.

The five mistakes that wreck a 13-week forecast

1. Including unsold pipeline as receipts. A 25%-confidence proposal is not cash. Track pipeline separately, or use a probability filter to view your forecast at multiple confidence levels.
2. Dating receipts by invoice issue date instead of expected payment date. Your receipts column should reflect the day money arrives in the bank, not the day the invoice was sent.
3. Lumping all expenses into "operating costs." The whole point of the forecast is to see specific cash events on specific days. A single line for "May expenses" hides timing.
4. Forecasting from accounting software exports. Accounting reports show what happened. The forecast shows what will happen. They are different documents and should not share a source of truth without manual review.
5. Refreshing monthly instead of weekly. A monthly refresh on a 13-week horizon is a quarterly exercise — exactly the cadence that hides the troughs the forecast was built to surface.

Adapting it to a service business

The classical 13-week forecast was built for industrial businesses with predictable AR, AP, and inventory cycles. Service businesses have a different shape:

• Receipts are lumpy and depend on individual clients, not aggregate demand.
• The biggest weekly disbursement is almost always payroll and contractor pay — and it's predictable to the cent.
• There is no inventory line, but there is often a deferred revenue line (retainers paid up front).
• Owner draws frequently exceed payroll and need a dedicated row.

For service businesses, the forecast also benefits from extending past 13 weeks — into a 52-week view — because seasonal patterns (summer slowdowns, year-end client budget freezes, Q1 ramp-ups) span more than a single quarter. The 13-week forecast is the discipline; the 52-week view is the strategy.

Getting started this week

A reasonable first build:

1. Open a spreadsheet. Thirteen columns across, dated by Monday of each week.
2. Top row: opening cash. Pull from your bank balance.
3. Middle rows: list every recurring receipt and disbursement, then drop them onto the right week.
4. Below that: list every open invoice on the date you actually expect payment.
5. Bottom row: ending cash, computed.
6. Schedule a recurring 30-minute Monday meeting with yourself.

That's the whole methodology. You can keep it in a spreadsheet forever — many operators do. The trade-off is that spreadsheets don't gracefully extend past thirteen weeks, don't model timing scenarios without copying tabs, and don't reconcile actuals vs. forecast without manual labor.

Runway Forecaster builds the same discipline into a 52-week visual grid: weekly receipts and disbursements, recurring frequencies, probability filters, and what-if scenarios. Try the demo — it's seeded with sample data so you can see what a working forecast looks like in five minutes. For the practical mechanics of building one, read the weekly cash-flow forecasting guide.