The single most useful financial discipline a small business can adopt is a 13-week rolling cash flow forecast, updated weekly. Done well, it catches most cash crunches with weeks of warning, makes timing decisions visible before they hit, and replaces gut-feel cash management with something you can actually plan against.

Definition

Cash flow forecast - a projection of expected cash inflows and outflows over a future period - week by week or month by month - that produces a running projected bank balance over time.

Why 13 weeks

Three months is the sweet spot for most businesses. Shorter and you can't see far enough to act on the warning; longer and the forecast accuracy degrades to the point where it's no longer trustworthy.

Businesses with longer cycles (construction projects, seasonal manufacturers) extend to 26 or 52 weeks at the cost of less accuracy at the far end. Businesses with short cycles (restaurants, fast-turn retail) sometimes shorten to 8 weeks. The principle is the same: forecast as far out as you can with usable accuracy.

What goes into a forecast

A useful forecast separates cash in from cash out, with enough line-item detail to identify what's driving each week.

Cash in

  • Receivables collections - existing invoices, scheduled by expected pay date (use historical pay patterns, not invoice terms)
  • Expected new sales - forecasted revenue that will become receivables, with realistic conversion timing
  • Recurring receipts - subscription revenue, contracts on retainer
  • Financing / other - planned loan draws, investor money, asset sales, tax refunds

Cash out

  • Payroll - by pay date, including payroll taxes
  • Recurring expenses - rent, utilities, software subscriptions, insurance
  • Vendor payables - by scheduled pay date
  • Debt service - loan principal + interest
  • Taxes - quarterly estimates, payroll taxes, sales taxes
  • Capital expenditures - scheduled equipment and asset purchases
  • Owner distributions - planned draws or dividends

Running balance

Starting cash + Cash in − Cash out = Ending cash. The next week starts from that ending balance.

Formula
Week N ending cash = Week N starting cash + Week N expected cash in − Week N expected cash out

Week N+1 starting cash = Week N ending cash

The receivables-timing problem

The single biggest source of forecast error is when customers will actually pay. Most owners forecast by invoice terms ("net-30 means they'll pay in 30 days"). Reality is messier - actual pay patterns vary by customer, industry, and economic environment.

Build your forecast from historical pay patterns:

  • For each customer (or customer segment), calculate average days from invoice to payment over the last 6-12 months.
  • Use that historical average for forecasting, not the invoice terms.
  • For your largest customers, build a per-customer pay schedule - they affect the forecast disproportionately.

A useful sanity check: total receivables ÷ average daily revenue ≈ Days Sales Outstanding. If your forecast assumes faster collection than your historical DSO, it's optimistic.

Make it rolling and update weekly

The forecast is alive, not a one-time exercise. Each week:

  1. Drop last week from the front.
  2. Add a new week at the back (so you always have 13 weeks ahead).
  3. Update actuals for what actually happened last week.
  4. Revise assumptions on anything you got materially wrong.
  5. Add new items that came in (a big invoice sent, a new contract signed, an unexpected bill).

The discipline of weekly update is what makes the forecast accurate over time. A perfectly-built forecast that gets revisited monthly is worse than a sloppy one revisited weekly.

Set thresholds and pre-decide actions

A forecast without a plan is just a number on a screen. For every forecast, set a clear threshold: "if my projected cash dips below $X at any point in the next 13 weeks, I act."

Pre-decide your actions in priority order:

  1. Aggressive receivables follow-up on anything past due
  2. Defer non-critical expenses (new hires, new equipment)
  3. Stretch payables to maximum without damaging relationships
  4. Draw on a credit line if one exists
  5. Have the financing-options conversation early, not late

Pre-deciding removes the "what do I do now?" paralysis that hits when cash dips suddenly. The forecast tells you when; the playbook tells you what.

Common forecasting mistakes

1. Building it once, never updating it

The most common failure mode. A four-month-old forecast is mostly fiction.

2. Forecasting receivables by invoice terms

Customers pay on their schedule, not yours. Use historical pay patterns.

3. Aggregating to monthly

Hides week-to-week pinch points. Payroll lands in specific weeks; quarterly taxes land in specific weeks. Monthly forecasts average them out invisibly.

4. Optimism on conversion timing

Deals signed today turn into receivables next month, not this week. Be honest about the lag.

Related concepts