A 13-week cash flow forecast template is a weekly grid of expected receipts and payments over one quarter, starting from a reconciled bank balance and ending each week with a closing cash figure and the headroom above a minimum you have set. The template on this page is a working spreadsheet: it pulls invoiced receipts from an open-invoice list by the date each customer is expected to pay, pulls supplier payments from a bill list by planned payment date, keeps financing separate from operating cash, and carries a variance sheet for the weeks already closed. The worked example inside it is a fictional 32-person engineering consultancy, with one week that dips below its cash floor, so you can follow every figure from ledger to closing balance.

Download the 13-week cash flow forecast template (Excel, 22 KB)

This page covers what is in the template and why, the worked example week by week, and the method behind the rows most free templates leave out. Whether to run a 13-week forecast, and the weekly rhythm around one, is in our guide to implementing a 13-week cash flow forecast; how far to trust the later weeks is in how accurate 13-week cash flow forecasts are.

What is in the template

Five sheets. The Forecast sheet is the grid: thirteen weekly columns, each headed with its Monday, and the rows below.

Opening cash is the reconciled balance across every bank account you include, excluding anything ring-fenced, at the Monday cut-off. Only the first week's figure is typed in; every later week opens with the previous week's closing balance.

Receipts come in three rows. Invoiced receipts are not typed into the grid at all: the Receivables sheet holds every open sales invoice with its due date and, separately, the date you expect the cash, and the grid sums those expected dates into weeks. Contracted receipts not yet invoiced are typed in as weekly estimates, timed by when the cash will arrive rather than when the invoice will be raised. Other receipts covers refunds, interest and the occasional one-off.

Operating payments follow the same split: approved supplier bills pulled from the Payables sheet by planned payment date (the Friday of the payment run, not the due date on the bill), supplier costs not yet billed as weekly estimates, then the rows that are regular but invisible in the invoice and bill flow: net pay, PAYE and National Insurance, pension, rent, software, insurance, VAT, corporation tax, capital expenditure, and interest and bank charges.

Financing sits beneath operating cash in its own block, the term loan principal and a line for a facility drawdown or repayment, so the grid shows the position before and after planned borrowing.

Closing cash is opening cash plus net operating cash flow plus net financing. Beneath it, a minimum cash floor you set once, a headroom row that subtracts the floor from each closing balance, and a flag for any week below the line. A small readout block then answers the questions the forecast exists to answer: the lowest closing balance in the window, which week it lands in, the first week below the floor, and how many weeks of headroom you have before it.

The Variance sheet holds the closed weeks: receipts and payments forecast against actual, kept apart so a miss on one side cannot hide a miss on the other, with a rolling four-week line beneath. Blue cells are inputs, black cells are formulas, and three yellow cells are set first: opening cash, the week 1 date and the floor.

The worked example: Hallam & Rowe Ltd

Hallam & Rowe is a fictional engineering consultancy in England with 32 staff, £4.2 million of annual turnover and Xero as its ledger. Every figure below is illustrative and internally consistent; nothing is drawn from a real business. The forecast runs from Monday 14 September to the week commencing 7 December 2026, with opening cash of £221,500 and a floor of £70,000, the sum the finance manager has decided the business must not drop beneath.

On the receipts side, the Receivables sheet carries twelve open invoices totalling £421,100. The largest is £61,200 from a water company, due 19 September and expected 25 September, because that customer pays about six days after due. The oldest is £42,600 from a rail systems customer, already overdue, expected on 16 September because the customer has confirmed its payment run. The one to watch is £58,000 from a local council, due 4 October and expected 21 October: in this example the council is the slowest payer on the ledger, and that expectation is set from its own payment history, not from the fact that it is a public body. From week 6 the invoiced receipts run out and the contracted work not yet invoiced takes over, with September's work landing at the end of October (£58,000, £112,000 and £84,000 in weeks 6 to 8) and October's at the end of November (£62,000, £124,000 and £96,000 in weeks 10 to 12). Receipts over the thirteen weeks total £1,038,300.

On the payments side, net pay of £108,500 leaves on the 25th, or the working day before, so it lands in weeks 2, 6 and 11. The PAYE and National Insurance for each pay run, £47,200, and the pension contributions, £9,800, leave on the 22nd of the following month, so August's fall in week 2, September's in week 6 and October's in week 10. Twelve approved supplier bills totalling £94,170 are paid in the Friday runs of weeks 1 to 5; after that, supplier costs are estimated at £22,000 to £27,000 a week. Rent of £27,000 is paid quarterly in advance on 29 September (week 3). The VAT return for the quarter to 31 August costs £68,300 on 7 October (week 4). Corporation tax for the year to 31 December 2025, £51,000, is due on 1 October (week 3). The insurance renews on 1 November for £14,500 (week 8). Survey equipment costing £18,000 is bought in week 7. A term loan takes £6,250 of principal and £1,450 of interest and charges on the 15th of each month. Operating payments total £990,020 and financing £18,750.

Here is the shape that produces, week by week.

The quarter ends £29,530 better than it started. Read only the first and last columns and the business looks comfortable. The row that matters is week 4.

Reading the result: the breach in week 4

Week 4 closes at £35,080, which is £34,920 below the floor. The readout block says it plainly: lowest closing cash £35,080, in week 4, the first and only week below the line, with three weeks of headroom before it. Nothing dramatic causes it. Corporation tax and rent leave in week 3, VAT leaves in week 4, and the receipts in those two weeks are the thin ones: £53,150 and £22,150, because the big September invoices were collected in weeks 1 and 2 and the council's £58,000 is not expected until week 6.

That is the value of a weekly grid over a monthly one. Weeks 4 to 7 taken together show receipts of £405,250 against operating payments of £339,530, a comfortable four weeks. The week-by-week view shows the business £34,920 short of its own floor for the seven days in which the VAT return is paid.

The lever is on the receivables sheet. If the council's £58,000 can be brought into week 4, by chasing the invoice against its 4 October due date rather than waiting for the customer's usual pattern, week 4 closes at £93,080 and the breach disappears. If it cannot, a £40,000 facility draw in week 4, repaid in week 5, does the same job: week 4 closes at £75,080 and week 5 is unchanged at £139,930. A third option is to move the week 4 supplier run of £10,630 into week 5, which narrows the gap but does not close it. The forecast does not choose between them. It tells you three weeks ahead that a choice is needed, and the size of it.

The grid also shows week 6, clearing the floor by £13,330, the narrowest margin after week 4, because it is a payroll week that also carries September's PAYE and pension. Two customers paying a week late would take it under; it is the week to protect once week 4 is handled.

How to use the template with your own figures

Step 1: Set the three yellow cells. Opening cash is your reconciled bank balance across every account you include, on the Monday you start. The week 1 date drives every column heading and every date lookup. The floor is the balance you have decided not to go below; the next payroll and tax run plus a margin is the usual starting point.

Step 2: Replace the Receivables and Payables rows with your own open invoices and approved bills. Leave each due date as the ledger has it, and set an expected date on every invoice from what you know: a confirmed payment run beats the customer's average, which beats the due date. For a bill, the planned payment date is the day your payment run will send it.

Step 3: Enter the recurring and statutory payments in the week the money leaves. Net pay on payday, employer taxes and pension on their statutory dates, VAT and corporation tax on their due dates, rent on the lease's date. Keep them on their own rows so a miss can be traced to a line.

Step 4: Estimate the receipts and supplier costs that have no invoice or bill yet. Time contracted work by when the cash will arrive, not when the invoice will be raised. Where you have no basis for a week, leave it empty rather than smoothing an annual figure across thirteen columns.

Step 5: Each week, close the completed week on the Variance sheet, then roll the grid. Record actual receipts and actual payments against the forecast, move the week 1 date forward by seven days, re-date the expected dates that have changed, retire any estimate that has become an invoice or a bill, and add the new week 13.

The method behind the rows

Expected dates, not due dates, drive the receipts. Two dates sit on every invoice in the Receivables sheet, and only one of them moves the forecast. A due date is what the contract says; an expected date is what the customer does. The template pulls invoiced receipts into weeks by expected date, and the worked example shows why: on due dates alone, the council's £58,000 would sit in week 4 and the breach would be invisible until the money failed to arrive. Most free 13-week templates in circulation carry a single date per invoice, if they hold invoices at all; the majority hold weekly totals typed straight into the grid, with the invoices and bills behind them somewhere else. That is the reason the weekly refresh becomes a data-assembly job rather than a review, and the reason the Receivables and Payables sheets exist here.

Invoiced and contracted receipts are separate rows for the same reason lenders test them separately: they are different grades of information. An invoice is a cash event with an uncertain date. Contracted work not yet invoiced is a date and an amount that both still depend on you. Keeping them apart means the first four weeks of the grid are built almost entirely from the first row, and any miss there is a ledger problem rather than a forecasting one.

The statutory rows carry their real dates, which in the UK means the following. PAYE and Class 1 National Insurance for a tax month (the 6th to the 5th) must reach HMRC by the 22nd of the next month if paid electronically, or the 19th by post, and when the 22nd is a weekend or bank holiday the payment must reach HMRC on the last working day before it, unless you pay by Faster Payments. In the example, the tax month from 6 October to 5 November is due on Sunday 22 November, so the payment is placed on Friday 20 November, in week 10. Employee pension contributions deducted from pay must reach the scheme by the 22nd of the following month at the latest under The Pensions Regulator's rules; your scheme may set an earlier date, and employer contributions follow the scheme's own schedule. A quarterly VAT return and its payment are due one calendar month and seven days after the period ends, so the quarter to 31 August is paid on 7 October; if you pay by Direct Debit, use the collection date rather than the deadline. Corporation tax is due nine months and one day after the accounting period ends for a company with taxable profits under £1.5 million, shared among associated companies, so a year to 31 December 2025 is paid on 1 October 2026. Rent is usually paid in advance, quarterly on the English quarter days (25 March, 24 June, 29 September and 25 December) or monthly; the lease decides, and the example assumes quarterly. None of these dates is moved by a bank holiday in the example's window: England and Wales have none between 14 September and 13 December 2026. Scotland has St Andrew's Day on Monday 30 November.

Financing is kept out of operating cash so the forecast answers two questions: whether the business generates the cash it needs, and what has been arranged for the weeks where it does not. Interest and bank charges sit with operating payments as a cost of running the business; the principal repayment sits with financing as the unwinding of a facility. Dividends, drawings or asset-sale proceeds get a row where they belong if you have them; the example has none.

The floor turns a column of balances into a decision. £35,080 means nothing on its own; £34,920 below the floor is a specific gap with a week attached, and the readout block does the arithmetic the board wants: how low, when, and how many weeks of warning.

The Variance sheet is where the forecast learns. In the example, four closed weeks show receipts £28,400 below forecast over the rolling four weeks, around 8% under, while payments are within 1%. That pattern says the expected dates on the receivables sheet are set too close to the due dates, and it says nothing about the payments side, which is exactly why the two are kept apart. The habit is: add the closed week, move the four-week window down a row, and carry what the variance taught you into next week's expected dates.

The week itself runs Monday to Sunday in this workbook. That is a design choice rather than a rule. Use whichever boundary your payroll, bank statements and payment runs are cut off on, and keep it.

Adapting the example outside the UK

The mechanics are the same in every market; the statutory rows change. In the United States, federal withholding and FICA deposits on the IRS monthly or semi-weekly schedule replace PAYE and National Insurance, and state sales tax replaces VAT. In Australia, superannuation must reach the fund within seven business days of each payday from 1 July 2026 under Payday Super, so the pension row becomes a per-payday row, and the quarterly BAS carries GST and PAYG withholding together. In New Zealand, PAYE and KiwiSaver deductions reach Inland Revenue by the 20th of the following month for most employers, and GST is paid on the 28th of the month after the period. Load your public holidays before fixing payment dates: Columbus Day, Veterans Day and Thanksgiving fall in this window in the US, Labour Day on 26 October in New Zealand, and state-specific holidays in Australia.

Public-sector customers deserve one note, because the example's slowest payer is a council. In the UK, public bodies are bound to pay a valid, undisputed invoice within 30 days of receiving it under section 68 of the Procurement Act 2023. The clock starts on receipt, so the date to forecast from is the date the invoice was accepted, not the date you sent it, and a disputed invoice stops the clock altogether. Whether a particular public customer pays inside that term is a question for its own payment history, which is where the example's expected date comes from.

How Float fits

Float is a cash flow forecasting tool that does what this template does, with the ledger doing the data assembly. It connects to Xero or QuickBooks Online through a one-way, read-only connection, imports reconciled bank transactions, invoices and bills once a day at an hour you choose, with a manual sync on demand, and offers both monthly and 13-week forecasting options using the direct method. Sage Intacct support is coming soon; you can join the waitlist.

The template's rows map onto Float's mechanics. The Receivables sheet's expected date is Float's expected payment date on every invoice and bill, set by hand, applied from each customer's own payment history, or split into part payments. The estimates for work not yet invoiced and costs not yet billed are budgets, which Float treats as placeholders that invoices, bills and transactions fill as they arrive. The minimum cash floor is a threshold line: set one and Float shows the date the balance is due to cross it. The weekly view shows the full 13-week forecast and exports it as a PDF or CSV from the Share button. What-if versions of the plan, such as the council paying in week 6 rather than week 4, are scenario layers on the base forecast, and every plan includes a set allowance of them, listed on our pricing page.

One thing to be clear about: this spreadsheet is a download from the blog, not a file that imports into Float. Budgets in Float are entered directly or pasted in from a spreadsheet grid, and the invoices, bills and bank balances arrive from your accounting platform. The template is the model on paper; Float is the model connected to the ledger.

Frequently asked questions

What is a 13-week cash flow forecast template?

A 13-week cash flow forecast template is a spreadsheet laid out as thirteen weekly columns that starts from a reconciled bank balance, schedules expected receipts and payments in the weeks the cash will actually move, and closes each week with a cash balance and the headroom above a minimum the business has set. The template on this page adds the invoice and bill lists behind the weekly totals, a separate financing block, and a variance sheet for the weeks already closed.

What rows should a 13-week cash flow forecast include?

No professional body prescribes a fixed row set; the guidance that exists from bodies such as ICAEW and the Association of Corporate Treasurers sets out principles and cash categories rather than a template. In practice the rows that matter are opening cash, invoiced and not-yet-invoiced receipts kept apart, payroll and its statutory follow-on payments on their real dates, supplier payments, the lumpy items such as VAT, corporation tax, rent and insurance, interest, financing shown separately, and closing cash against a floor. Add rows for dividends, drawings or asset sales only where they are material.

Why does the template use expected payment dates instead of due dates?

Because customers pay on their own timetable, not the contract's. A forecast built on due dates places every receipt too early and hides the weeks where cash is short, which in the worked example would have made the week 4 breach invisible. The expected date on each invoice comes from a confirmed promise, the customer's known approval cycle, or its average payment behaviour, in that order of preference.

How do I set the minimum cash floor?

Start from the cash you need to make the next payroll and tax run with a margin for two customers paying late, and adjust it as the variance sheet shows how far your receipts drift. In the worked example the floor is £70,000 against a monthly net payroll of £108,500, which is a deliberately tight setting chosen to show what a breach looks like; most businesses at this size would set it higher.

What should I do when the forecast shows a week below the floor?

Read three numbers first: how far below, which week, and how many weeks of warning. Then work the levers in order of cost: bring a receivable forward by chasing it against its due date, move a discretionary payment such as a supplier run or a capital purchase back a week, and only then plan a facility draw. In the worked example, securing one council invoice a fortnight earlier clears a £34,920 gap; a £40,000 draw repaid the following week does the same.

How do I roll a 13-week forecast forward each week?

Close the completed week on the Variance sheet by recording actual receipts and payments against the forecast, move the week 1 date forward by seven days, re-date any expected dates that have changed, replace estimates that have become invoices or bills, and add the new week 13. The whole cycle, including a short review of overdue receivables, takes most finance teams at this size under an hour once the lists are maintained.

Is the 13-week forecast a direct-method forecast?

This template uses the direct method: receipts and payments scheduled in the weeks the cash moves, the same meaning the accounting standards give the term for a cash flow statement, rather than profit adjusted for non-cash items. Short-term forecasting is usually built this way because the question it answers is when money moves, but other methods exist, and the 13-week horizon does not by itself require one.

Does a Scottish or non-UK reader need to change anything?

Only the statutory rows and the holiday calendar. The example is set in England, where no bank holiday falls in its window; Scotland has St Andrew's Day on 30 November 2026. Readers in the US, Australia and New Zealand replace PAYE, pension and VAT with their own payroll tax deposits, superannuation or KiwiSaver, and sales tax or GST, on their own dates, and load their public holidays before fixing payment dates.

Can I import this template into Float?

No. Float builds its forecast from the invoices, bills and bank balances in Xero or QuickBooks Online, and budgets are entered directly or pasted in from a spreadsheet grid. The template is a model to run by hand or to compare against; if you connect your accounting platform to Float, the invoiced-receipts and approved-bills rows arrive on their own and the expected dates, budgets and threshold do the rest.

Who can see the forecast in Float?

Float uses three roles: Admin, Editor and Viewer. The finance team keeps edit rights, leadership and the board can be given read-only access, and the connection to the accounting platform is one-way, so nothing done in Float changes the books. Two-factor authentication is mandatory for Xero users.

The template gives you the model on paper, with a quarter of figures to check your own against. Connect Xero or QuickBooks Online to Float and the invoice and bill lists build themselves, with a 14-day free trial and about three minutes to your first 13-week view.