Startup finance

How to build a weekly cash flow forecast for a startup

A 13-week cash flow forecast is a weekly schedule of the cash you expect to receive and pay over the next quarter, starting from the balance in your bank account today. It follows the day money moves, not the day a sale or bill is booked. Its job is to show the lowest cash point of the quarter while you can still change something.

What a 13-week cash flow forecast is

Thirteen weeks is one quarter. That is far enough ahead to see a payroll crunch or a slow-paying customer coming, and close enough that you can fill most lines from real invoices, bank data and the payroll calendar instead of guesses.

The forecast uses the direct method. You list actual receipts and disbursements (the cash paid out) by week, rather than starting from profit and adjusting for accruals. Guides from Atlar and others note that longer forecasts usually switch to the indirect method, which is why this one is paired with a longer plan and does not replace it.

Each week follows one formula:

ending cash = opening cash + receipts - payments

The ending cash of week 1 becomes the opening cash of week 2, so a change in any week flows through every week after it. A single late payment visibly lowers every later balance, and that chain is what makes the forecast useful.

Why startups use it

A startup that is burning cash lives close to the edge between “plenty of time” and “payroll is at risk”. A monthly budget can hide the problem, because three good weeks average out one dangerous week, and a weekly view shows that week.

Typical moments when the forecast earns its keep:

  • Before a fundraise. Investors want to know how long the money lasts, and a forecast tested against actuals each week is more convincing than a runway guess. See three steps to getting funded for what else is asked at that stage.
  • When runway is short. If the balance dips below a safe level, you want to learn it in week 4, not in week 11.
  • When collections are lumpy. One large invoice that pays late can matter more than a month of small ones.
  • When a lender or board asks for it. Some credit agreements and boards request a short-term cash view as a routine report.

Graphite Financial adds that this tool suits businesses that grow fast, burn cash or carry inventory, and that it is a near-term view, not a stand-in for a long-range financial model. Burn rate and runway explains the longer-horizon side.

What goes into the layout

Most templates share the same skeleton: one row per cash line, one column per week. Dating each column by its week-ending date (for example “Fri Dec 5”) is clearer than “Week 1”, because the dates line up with payroll runs and rent days.

Section Typical lines
Opening position Opening bank balance, minimum cash you want to keep
Receipts Customer collections from open invoices, new bookings paid upfront, renewals, interest, grants and tax refunds, financing (kept on its own line)
Payments Payroll and payroll taxes, contractors, rent, software and cloud bills, marketing, insurance, professional fees, taxes, debt repayments
Summary Net cash flow, ending cash, cash above or below the minimum, weeks of runway

Two design choices matter more than the rest:

  1. Keep financing on its own line. A loan draw or equity wire is not operating cash, and mixing it into receipts makes the business look healthier than it is.
  2. Do not net lines. Show receipts and payments gross, since netting hides which side is moving.

Six steps to build it

The steps below combine the approach in the guides from Graphite Financial, Decipher and CFO Advisors.

1. Start with a cash number you trust

Take today’s balance from the bank and reconcile it to your books. Subtract money that is not really available, such as checks that have not cleared or funds earmarked for tax. Then set a minimum cash level: the amount below which you would act. A common starting point is the cash needed to cover a couple of payroll cycles. CFO Advisors presents its own version of that rule as a recommendation, not a standard, so set the floor that fits your risk.

2. List expected receipts by week

Start from open invoices and add the typical number of days your customers pay late to each due date. A customer that pays 12 days after the due date should appear in that week, not on the due date. Then add receipts without an invoice, such as card or marketplace payouts and new prepaid deals. Mark each line as confirmed or uncertain. Revenue that is booked but not collected has no place in this table.

3. List expected payments by week

Payroll comes first, because it is usually the largest and least flexible line. Enter every payment in the week it clears the bank, not the week the bill arrives. Add rent, software, contractors, insurance, taxes and loan payments. Ask the people who control spend, such as the head of sales or engineering, what they plan to commit this quarter. A forecast built by finance alone tends to miss things.

4. Calculate ending cash and find the low point

Apply the formula to every week and scan the ending-cash row. Ask when the lowest balance falls and whether it stays above the minimum, since the quarter-end figure matters less.

5. Test what could go wrong

Pick the two or three assumptions that move cash most, usually a large customer paying late or a hire starting earlier than planned. Re-run the sheet with a slip. Graphite’s own example shows week 3 ending cash falling from $78,000 to $38,000 when a $40,000 payment arrives one week later. Decipher’s sample does something similar with a two-week collection delay that pulls the first week below the minimum forward from week 9 to week 5.

6. Name an owner for every line

A line without an owner stops being updated. Write next to each row who supplies the number and where it comes from (billing system, payroll provider, bank feed).

Worked example

The table below is made up for illustration. It describes a small software company that starts the quarter with $300,000, collects about $18,000 a week from subscriptions with a few larger invoices, pays $40,000 of payroll every second week, and has rent and other bills that vary. All numbers are in thousands of US dollars.

Week Opening cash Receipts Payroll Other payments Net flow Ending cash
1 300 18 0 14 +4 304
2 304 16 40 9 -33 271
3 271 35 0 7 +28 299
4 299 15 40 10 -35 264
5 264 18 0 14 +4 268
6 268 20 40 8 -28 240
7 240 22 0 7 +15 255
8 255 16 40 12 -36 219
9 219 18 0 14 +4 223
10 223 30 40 8 -18 205
11 205 18 0 7 +11 216
12 216 20 40 9 -29 187
13 187 18 0 14 +4 191

What the sheet says:

  • Cash falls from $300,000 to $191,000, a net burn of $109,000 over 13 weeks, about $8,400 a week.
  • At that average pace, $191,000 lasts roughly 23 more weeks. This is the runway line many templates add: ending cash divided by the recent average weekly net burn.
  • The pattern is a sawtooth. Weeks with payroll lose $28,000 to $36,000, and the quiet weeks in between recover a little. A monthly view would smooth all of it away.
  • The $35,000 invoice in week 3 matters. If it slipped past week 13, ending cash would be $156,000 and week 4 would end at $229,000 instead of $264,000. Nothing breaks in this case, but the margin is thinner than the headline suggests.

Replace the invented numbers with your own, and keep the structure.

How to keep it current

A forecast is only worth the effort if it rolls. A rolling forecast always shows the next 13 weeks, so the horizon moves forward every week and the model never goes stale. The usual routine, described by Upflow and Whipplewood, takes under an hour:

  1. Each Monday, enter last week’s actual receipts and payments.
  2. Replace the forecast figures for that week with the actuals and compare them.
  3. Add a new week at the end so the sheet always covers 13 weeks.
  4. Revise the assumptions that missed, for example the average days late for a customer.
  5. Re-check the lowest balance and the minimum-cash line.

Track the gap between forecast and actual for each line. Decipher suggests a monthly tie-out to the budget as well. CFO Advisors proposes grades for the size of the variance (under 5% excellent, over 20% a red flag), which it describes as its own guideline, so treat them as a starting point. Upflow notes that there is no widely accepted independent benchmark for forecast accuracy at mid-market scale, so the useful measure is whether your own error shrinks over time.

Common mistakes

  • Forecasting revenue instead of collections. A sale in March that is paid in May belongs in May.
  • Leaving out known payments. Annual software renewals, insurance premiums and tax installments are easy to forget and easy to schedule.
  • Building it once. An outdated forecast is worse than none because it gives false comfort, and refreshing it only monthly defeats the point.
  • Mixing financing into operating receipts, which hides the real burn.
  • No link to the longer plan. The 13-week view should agree with the annual budget. A financial model that people trust and a financial dashboard work together with it.

For a wider list of money errors that show up in early companies, see top finance mistakes.

Questions people ask

Does a startup need a special tool? No. A spreadsheet is enough at the start. Several vendors offer free templates, and all of them are the same grid described above. Check any formulas before relying on them, since many templates come from companies that sell related services.

How is it different from a cash flow statement? A cash flow statement reports what happened over a past period, usually after the books are closed. The 13-week forecast looks forward and is updated weekly.

Is 13 weeks always the right horizon? It is a convention. Some teams use 8 or 26 weeks. What matters is that the horizon is short enough to fill with real data.

Who should own it? Usually whoever runs finance: the CFO, a controller or the founder in a small team. Inputs come from sales, payroll and operations.

Sources and limits

  • Graphite Financial, “What is a 13-Week Cash Flow Forecast? A Practical Guide For Startups” (accessed November 2026): formula, six build steps, week 3 example.
  • Decipher, “13-Week Cash Flow Forecast: How to Build the Model” (accessed November 2026): seven steps, sample company, monthly tie-out.
  • CFO Advisors, “13-Week Cash Flow Forecast for Startups: 2026 Template and CFO Walkthrough” (accessed November 2026): template lines, variance grades, minimum-cash rule.
  • Upflow, Whipplewood, Atlar and Stampli guides (accessed November 2026): rolling routine, direct versus indirect method, accuracy.

Most sources are vendor blogs, so they are practical guidance and not standards. The worked example uses invented numbers and is not data from any company. Figures such as variance thresholds and minimum-cash levels are suggestions from those vendors, and your own situation may call for different values.