Skip to content
The Insight Journal Business & Operations

Finance & Cash Flow · Method

How to build a cash flow forecast that shows you the gap five weeks early

A cash flow forecast puts a number on a week that has not happened yet. Every page ranking for this tells you to work out your running balance, then stops. This one runs the arithmetic across eight weeks and shows a month that closed at $2,000 with a week inside it that was $9,100 short.

Reviewed August 2026 · The Insight Journal Editorial Team

In short

A cash flow forecast is a forward-looking schedule of the money expected to enter and leave a bank account, arranged week by week, so the closing balance of every future week can be read before it arrives. Its output is a date and an amount. That is what separates it from a budget, which reports a period.
Building a cash flow forecast: a bookkeeper works between a laptop and a printed weekly schedule at a home office desk.

Scope

What a cash flow forecast is, and the one thing it is for

In short

A cash flow forecast estimates the cash position of a business at a series of future dates. It works by carrying a running balance forward: opening balance, plus receipts, minus payments, equals the closing balance, which becomes the next period's opening balance. Built weekly, it identifies the specific week a shortfall lands rather than the month it belongs to.

The calculation is trivial. Anyone can add and subtract, and the formula fits in one line, which is why every competing page can publish it and still leave the reader unable to build one.

The difficulty is entirely in the inputs. Each figure has to come out of a real document with a real date on it, and most of this page is about which document, and which date.

What a forecast is not: a profitability statement, a plan, or a target. It contains no opinion about whether the business is any good. It only answers whether the money will be there.

Everything upstream of this page, meaning what cash flow management is, where the operating, investing and financing categories come from, and how the accounting method changes what your books can show you, sits in the wider guide to cash flow management. This page starts where that one stops, at the sheet itself.

909

Words in the PwC page holding a page-one slot for "how to build a cash flow forecast"

Measured live via DataForSEO, 18 August 2026

0

US government or standards sources ranking on page one of either query

Live search result pulls, 18 August 2026

2

Public-interest results serving US searchers, one Australian and one British

Live search result pulls, 18 August 2026

1,900

US monthly search demand for "cash flow forecasting", trending 33% down year on year

DataForSEO, 18 August 2026

The horizon

Why thirteen weeks, and why the rows are weeks

Google's People Also Ask box asked this on both queries we pulled: "Why is a cash flow forecast 13 weeks?" and "Why 13 week cash flow?". Every page that uses the number states it as received wisdom and moves on.

In short

Thirteen weeks is one quarter, and a quarter is the shortest horizon that reliably contains the events that cause shortfalls: several payroll dates, one of the four dated estimated tax payment periods, every monthly contract renewal, and usually a seasonal turn. It is also short enough that most inputs are still documents rather than assumptions.
Weeks 1 to 2 Cleared and committed Documents only Weeks 3 to 6 Invoiced, on terms High confidence Weeks 7 to 10 Contracted, not billed Timing risk Weeks 11 to 13 Expected activity Assumption row one quarter, re-cut every week so the horizon never shortens confidence falls left to right, and the sheet should label the far end as assumption rather than pretend otherwise
The four confidence bands are an editorial framing of a standard rolling horizon, not a threshold published by anybody. Sources for the events the quarter contains: IRS estimated tax payment periods.

What a quarter is long enough to contain

The IRS states that "for estimated tax purposes, the year is divided into four payment periods," each with its own due date. A thirteen-week horizon therefore crosses at least one of them, which a four-week or eight-week view can miss entirely.

It also contains six fortnightly payrolls or three monthly ones, every monthly contract renewal you hold, and for most seasonal businesses one turn of the season. Those four categories cover the large majority of shortfalls.

Past about a quarter, the inputs stop being documents. A week thirty-five figure is a guess wearing a spreadsheet cell, and a guess in a cash forecast is worse than an empty row because it looks like information.

None of this makes thirteen a rule, and no authority publishes it as one. It is a working default, and a firm with weekly payroll and thirty-day terms could reasonably run ten.

Rolling means one thing

Every week you add a new week thirteen. The horizon never shortens, so you are never looking at a five-week view in the fifth week of the quarter, which is precisely when a business stops seeing the next tax date.

Why the rows are weeks

Monthly rows net a payroll date against a receipt that arrives eleven days later and report a comfortable month. The arithmetic proving it sits two sections down, and it takes four numbers.

Under ten people

None of this happens in a finance function at that size; it happens on somebody's Sunday evening. That is a different job, taken apart in the version written for a team of under ten people.

The method

Direct or indirect, and why the answer is settled for an operator

Page one lists both methods and decides between neither. The longest page we measured runs a heading for each and leaves the reader to pick, which is not much help to somebody with a bank balance and a payroll date.

In short

Use the direct method. It builds the forecast from actual receipts and payments, dated, which is the only way to produce a closing balance for a specific week. ASC 230 encourages entities to report operating cash flows directly, by showing major classes of operating cash receipts and payments, and stops short of requiring it.
Three ways of producing a cash flow number, what each starts from, what each outputs, and who each is for
Method Starts from Produces Who it is for
Direct Actual receipts and payments, dated A closing bank balance for every future week Anyone who needs to know whether Friday works
Indirect Net income, adjusted for non-cash items An explanation of why cash moved last period Reporting, lenders, and backward-looking analysis
Three-way A linked profit, balance sheet and cash model A projected balance sheet alongside the cash line Fundraising, acquisitions, covenant testing

Why the indirect method cannot answer the question

Under the indirect method, as BDO's explainer puts it, the entity adjusts net income for revenue and expense items that were not the result of operating cash transactions during the reporting period. It starts from a period result and works backwards.

That is a good way to explain why cash moved last quarter, and it is genuinely useful to a lender reading your accounts. It cannot tell you about Friday, because net income has no dates in it.

One ranking page on this query derives the forecast from the balance sheet and income statement inside a financial model. That is not a worse answer, it is an answer to a different question, and a reader who follows it looking for a payroll date will not find one.

There is also no conflict to resolve here. A business can report indirectly to its lender once a year and forecast directly for itself every week, and most do.

The three-way forecast, scoped

A three-way forecast links the profit and loss, the balance sheet and the cash flow so a change in one updates the others. Google's People Also Ask box asks what it is; page one mostly answers with a feature list.

It is a financial model, not an operating tool. It needs a reliable balance sheet underneath it, it earns its cost during fundraising or covenant testing, and it will not tell you anything about next Tuesday that a direct weekly sheet does not.

Run the weekly direct forecast for a full quarter before anyone spends money building a three-way model.

The inputs

Where every number in the forecast actually comes from

"List all your income. List all your outgoings." That is the instruction on the page-one result we measured at 909 words, and it is the instruction on all of them. None names the document a single line is read out of.

In short

Every line in a cash flow forecast is read out of a specific document, never estimated. The opening balance comes from the reconciled bank statement, receipts from the aged receivables report adjusted for observed payment behaviour, payroll from the register by pay date, and payroll taxes from the federal deposit schedule that applies to you.
Nine forecast input lines, the document each is read from, which date to enter, and the mistake most commonly made on each line
Line Read it from Which date to enter The common mistake
Opening balance The reconciled bank statement Cleared funds only, minus anything already committed by card or standing payment Using the ledger balance instead, which includes money that has not moved
Customer receipts The aged receivables report The invoice date plus the agreed terms, then adjusted for how that customer actually pays Entering the terms date rather than the observed date for a customer who is always eleven days late
New sales The pipeline, clearly marked Only work that is contracted; anything else belongs in a labelled assumption row Mixing hoped-for revenue into the same column as invoiced revenue
Payroll The payroll register The pay date, not the period the work covered Averaging a fortnightly payroll into a monthly figure
Payroll taxes The IRS deposit schedule that applies to you The deposit date set by federal rule, which can fall in a different week from the payroll Assuming the tax leaves the account on payday
Rent, insurance, subscriptions The contracts The billing date written in the agreement, including annual renewals Forgetting the once-a-year renewals entirely
Loan repayments The amortisation schedule The debit date, at the full amount including interest Treating a repayment as a cost rather than as a dated cash movement
Supplier payments The aged payables report The date you have actually agreed to pay, not the date on the invoice Paying earlier than agreed out of habit, which quietly funds the supplier
Owner draws The owner Every planned withdrawal, on its date Leaving them out because they do not feel like a business expense

The outflow whose date you do not choose

Payroll taxes land on a federal schedule

IRS Publication 15 sets two deposit schedules for employment taxes. Monthly schedule depositors deposit by the 15th of the following month.

For semiweekly depositors the publication states that liabilities incurred on Wednesday, Thursday and Friday are due the following Wednesday, and liabilities incurred on Saturday, Sunday, Monday and Tuesday are due the following Friday.

Which schedule applies is set by a lookback period, described in the publication as the four quarters of the preceding year, July 1 through June 30. Separately, if accumulated undeposited employment taxes reach $100,000 or more on any day in a deposit period, the deposit is due the next business day.

The forecasting consequence is the whole point of quoting it: a payroll and the tax on that payroll can land in different weeks. That is invisible in a monthly view and material in a weekly one.

Confirm which schedule applies to you with your accountant or payroll provider, not with a web page. Nothing here is tax advice.

Two rules that do most of the work

First, enter the date the money moves, never the date the paperwork was raised. An invoice issued today on net 30 terms to a customer who has paid on day 41 for six consecutive invoices belongs in week six.

Second, keep confirmed items and assumptions in visually separate blocks. Anything whose date you are guessing goes in a marked assumption row, so nobody reads a hope as a fact three weeks later.

The aged receivables report is the single most valuable document in this list, because it is the only one that records what customers actually do rather than what they agreed to do.

If producing these reports is itself the obstacle, that is a ledger problem rather than a forecasting one, and it starts with choosing small business accounting software. Whether your books can even surface the problem is covered in how the accounting method changes what your books can show you.

Worked example

Eight weeks of a seasonal contractor, with the arithmetic shown

These are illustrative round figures for a hypothetical seasonal contractor coming off a busy season. They are arithmetic, not a cited statistic, not data about any real business, and not a benchmark for yours. The method is the transferable part.

In short

Opening balance $38,000. Fortnightly payroll of $22,000, payroll tax deposits of $6,400 landing the week after each payroll, a quarterly estimated tax payment of $9,500 in week six, plus rent, materials, insurance, an equipment loan and owner draws. The trough is week six at minus $9,100, and it is visible in week one.
Eight weeks of a thirteen-week rolling cash flow forecast on illustrative hypothetical figures, showing cash in, cash out, what lands each week, net movement and closing balance
Week Cash in Cash out What lands Net Closing
Opening Reconciled bank balance, cleared funds only $38,000
Week 1 $26,000 $10,500 Rent 4,500 · materials 3,000 · owner draw 3,000 +15,500 $53,500
Week 2 $9,000 $29,300 Payroll 22,000 · materials 5,500 · equipment loan 1,800 -20,300 $33,200
Week 3 $21,000 $8,400 Payroll tax deposit 6,400 · materials 2,000 +12,600 $45,800
Week 4 $7,000 $27,200 Payroll 22,000 · materials 4,000 · insurance 1,200 -20,200 $25,600
Week 5 $12,000 $16,900 Payroll tax deposit 6,400 · rent 4,500 · materials 3,000 · owner draw 3,000 -4,900 $20,700
Week 6 $6,000 $35,800 Payroll 22,000 · materials 2,500 · equipment loan 1,800 · estimated tax 9,500 -29,800 -$9,100
Week 7 $24,000 $9,900 Payroll tax deposit 6,400 · materials 3,500 +14,100 $5,000
Week 8 $21,000 $24,000 Payroll 22,000 · materials 2,000 -3,000 $2,000

Illustrative arithmetic on hypothetical round figures. Weeks nine to thirteen are omitted for legibility; in practice they exist and carry the assumption rows. Every closing balance is the previous closing plus cash in minus cash out.

Read every week, not the last one

Week eight closes at $2,000, which is positive. Reading only the end of the horizon would report a business that finishes the period solvent and miss the week it was not.

Why week six breaks

Three dated outflows collide: a $22,000 payroll, an $1,800 loan repayment and a $9,500 estimated tax payment, against $6,000 of receipts in the quietest week of the season.

The tax lag is doing real work

Payroll taxes of $6,400 land in weeks three, five and seven, one week after each payroll. Entered on payday instead, they would shift $6,400 into week six and take the trough past minus $15,000.

The proof

The month that closed at $2,000 while week six was $9,100 short

A page-one result on the how-to query tells the reader to create twelve columns across a spreadsheet representing the next twelve months. Here is what that resolution does to the same eight weeks.

In short

Take weeks five to eight as a single month. Inflows total $63,000 and outflows total $86,600, a net movement of minus $23,600 against an opening balance of $25,600. The month closes at $2,000, positive. Inside it, week six closed at minus $9,100.

The same four weeks, both ways

  1. 1. Opening balance entering week five: $25,600.
  2. 2. Inflows across weeks five to eight: 12,000 plus 6,000 plus 24,000 plus 21,000, which is $63,000.
  3. 3. Outflows across the same four weeks: 16,900 plus 35,800 plus 9,900 plus 24,000, which is $86,600.
  4. 4. Net movement: 63,000 minus 86,600, which is minus $23,600.
  5. 5. Closing balance: 25,600 minus 23,600, which is $2,000. The month is positive.

Illustrative arithmetic on the hypothetical figures in the table above, not a cited statistic.

What the monthly row does not say

The monthly row reports $2,000 and stops. It is arithmetically correct and operationally useless, because the business could not have reached the end of that month without covering week six.

Netting is the mechanism. A month is a container, and inside the container a $22,000 payroll and a $9,500 tax payment in week six are cancelled out by a $24,000 receipt in week seven that had not arrived yet.

That is the entire case for weekly rows, and it takes four numbers rather than an argument. It is also why a twelve-month monthly view is a budget: useful for planning the year, incapable of answering a solvency question about a Friday.

Keep both if you like. Just never let the budget be the thing you check before committing to a payment.

The payoff

Closing the gap, and how much of it is real money

A forecast that identifies a trough and changes nothing is a diary. This is the part where five weeks of notice turns into a different week six, and where it pays to be honest about which moves earned money and which only moved it.

In short

Three moves take week six from minus $9,100 to plus $7,400: collecting $6,000 of overdue invoices, taking $8,000 of deposits on work scheduled for week seven, and moving a $2,500 materials payment out to week eight. Only the first is new money. The eight-week closing balance rises from $2,000 to $8,000, and that $6,000 improvement is the receivable.

  1. Collect the two oldest overdue invoices

    Two invoices totalling $6,000 have sat past 45 days. Chased in week one and paid in week five, they are the only move here that brings in money the business did not already have coming.

    +$6,000 of new cash

  2. Take deposits on the week seven work

    Two jobs scheduled for week seven are worth $8,000 in deposits if billed at the point of scheduling. That money arrives in week five, which means week seven arrives $8,000 lighter. It is borrowed, not earned.

    $8,000 moved earlier

  3. Move the materials order out two weeks

    The supplier has agreed to net 30 rather than net 15 on a $2,500 order. The cost is unchanged, the goodwill is spent once, and the payment lands in week eight instead of week six.

    $2,500 moved later

The revised weeks, worked through

  1. 1. Week five gains $14,000, being $6,000 collected and $8,000 in deposits, so it closes at $34,700.
  2. 2. Week six outflows fall by $2,500 to $33,300, against $6,000 in, so the net is minus $27,300.
  3. 3. Week six now closes at 34,700 minus 27,300, which is $7,400 rather than minus $9,100.
  4. 4. Week seven arrives $8,000 lighter, so it closes at $13,500 instead of $5,000.
  5. 5. Week eight carries the deferred $2,500, so it closes at $8,000 instead of $2,000.

Illustrative arithmetic on the hypothetical figures above, not a cited statistic and not a recommendation.

New money against moved money

Over eight weeks the closing balance improved by exactly $6,000, which is exactly the overdue invoice value. The other $10,500 of activity changed nothing about how much the business earned.

That is not a criticism of the other two moves. Moved money is what kept week six solvent, and solvency in week six is the only reason weeks seven and eight exist.

It matters because moved money is borrowed from your own future weeks, and doing it every quarter without ever fixing collections produces a business that is permanently four weeks ahead of itself.

The wider set of tactics, with each one costed, sits in the moves available to an operator on thin margins. For a product business the largest reversible number is usually stock, which is where inventory and supplier timing stops being a logistics question.

The review

Reviewing the forecast against what actually happened

"Compare actuals to forecasts" is the closing instruction on almost every page we read. Nobody says what to do with the difference, which is the step that decides whether the sheet ever gets better.

In short

Variance comes in three types and each carries a different fix. A timing variance means the amount was right and the date was wrong. An amount variance means an assumption is wrong and should be rebuilt. An omission means a whole line was missing and will keep being missing until it is added permanently.
Three types of forecast variance, what each means, a typical example, and the fix for each
Type What it means Typical example The fix
Timing The amount was right and the date was wrong A customer paid on day 34 against 30-day terms Adjust that customer's observed payment lag, not the invoice amount
Amount The date was right and the number was wrong A materials order came in 18% over the estimate Rebuild the assumption behind the line, because it will be wrong again next week
Omission A whole line was never in the sheet An annual software renewal debited without warning Add it permanently and check the contracts folder for its siblings

The thirty-minute weekly loop

  1. 1. Reconcile the bank and take the real opening balance.
  2. 2. Enter what actually happened last week beside what you predicted.
  3. 3. Type each difference as timing, amount or omission.
  4. 4. Apply the fix to the input, not to the closing balance.
  5. 5. Extend the horizon by one week so it stays at thirteen.
  6. 6. Read every closing balance in the horizon, not just the last.

Step three is the one that gets skipped, and skipping it is why so many forecasts are wrong in the same direction for a year without anybody noticing the pattern.

Bias beats accuracy

A forecast that is wrong by a similar amount in both directions is healthy. A forecast that is optimistic every single week has a systematic fault, usually receipts entered on terms dates instead of observed ones.

Never fix the answer

Overwriting a closing balance to match the bank hides the input that was wrong. Correct the input line and let the closing balance recalculate, or the sheet stops teaching you anything.

Where this belongs

The weekly loop is a reporting cadence, not a finance task, and it works best sitting inside the operations function this reporting sits inside rather than in a private spreadsheet nobody else can open.

The boundary

When the forecast stops being the right tool

In short

A forecast is a measuring instrument, and there is a point where measuring more carefully stops helping. When the sheet says the gap does not close, when shortfalls repeat across quarters, or when tax obligations are involved, the next step is a certified public accountant or an SBA resource partner rather than another tactic.

Signals the sheet has done its job

  • The trough reappears every quarter regardless of which levers are pulled.
  • Closing the gap now depends on moving money that has already been moved once.
  • Obligations to a tax authority are part of the shortfall.
  • The horizon extends and the negative weeks extend with it.
  • Two people run the same sheet and get different answers.

Where to take it

The SBA's guidance on managing business finances notes that a CPA typically costs more than online services but can normally offer more tailored service for your specific business needs, while a bookkeeper provides basic day-to-day functions at a lower cost without the formal accounting education.

The same guidance points readers to free business counseling and to finding counselors locally, which costs nothing and is a reasonable first call before a paid engagement.

If the underlying pattern is a business growing faster than it collects, that is a structural problem a longer horizon will not solve, and it is the subject of growth that outruns its own collections cycle.

Nothing on this page is financial, tax or legal advice. It describes how to build and read a forecast, and it deliberately does not tell you which obligation to satisfy, which product to buy, or what your tax position is.

Method

How we researched this page

Measured, not remembered

Two live search result pulls on 18 August 2026, US English, desktop, on "cash flow forecasting" and on how to build one. Competitor lengths were measured rather than estimated: 3,066 words for the longest page on the search result, 2,014 for a fractional-CFO firm, and 909 words and four steps for the PwC post holding a page-one slot.

The finding worth repeating

No United States government or standards source ranks on page one of either query. The two public-interest results serving American searchers are the State Government of Victoria in Australia and a UK charity umbrella body writing for charity trustees.

A community forum thread outranks every publisher on the how-to query. It blocked our crawler, so it is described only from its live snippet and never from memory.

What we left out

No estimated tax calendar due dates appear here. The IRS page we read referred out to Form 1040-ES rather than listing them, so we did not verify them, and no deposit-schedule dollar threshold appears for the same reason.

No software name, price, rating, interest rate or template download appears anywhere on this page. We sell no forecasting product and took no placement. Our sourcing and disclosure standard sets out the rest.

Questions

Common questions about cash flow forecasting

Why is a cash flow forecast 13 weeks?
Because thirteen weeks is one quarter, and a quarter is the shortest window that reliably contains the events that cause shortfalls. In that span you will normally cross at least six fortnightly payrolls or three monthly ones, one of the four dated estimated tax payment periods the IRS sets out, a full billing cycle on every monthly contract you hold, and at least one seasonal turn. It is also short enough that most of the inputs are documents rather than guesses, which is what separates a forecast from a budget. Nothing about the number is a rule, and no authority publishes it as one. It is a working default that survives because a shorter horizon misses the tax date and a longer one fills up with assumptions.
How do you calculate a cash flow forecast?
For each week: closing balance equals opening balance, plus cash received that week, minus cash paid that week. The closing balance then becomes the next week's opening balance, and you repeat across the horizon. That is the whole calculation, and its difficulty is not arithmetic but sourcing: every figure has to be read out of a document rather than estimated. The worked example on this page runs the calculation across eight weeks, starting from an opening balance of $38,000 and finishing at $2,000, with a trough of minus $9,100 in week six. Those figures are illustrative arithmetic on hypothetical round numbers, not data about any real business.
What is a 3-way cash flow forecast?
A three-way forecast links three statements so they move together: the profit and loss, the balance sheet and the cash flow. Change an assumption in one and the other two update, which is what makes it useful for testing a covenant, modelling a funding round or valuing a business. It is a financial model rather than an operating tool, and it needs a reliable balance sheet to sit on. Most owner-run businesses do not need one and should not build one first. Get a weekly direct forecast working for a full quarter before anyone spends money on a three-way model.
How do I do a 12 month cash flow forecast?
You can, and it answers a different question. A twelve-month view with monthly columns is a budget: it tells you whether the year works, which is a planning question. A thirteen-week view with weekly rows is a forecast: it tells you whether a specific Friday works, which is a solvency question. Many businesses keep both and never confuse them. The failure mode is running only the twelve-month version, because monthly columns net a payroll date against a receipt eleven days later and report a comfortable month that contained a week with no money in it.
What are the four types of forecasting?
There is no settled set of four in cash forecasting, and the question is worth answering plainly rather than inventing one. The distinctions that actually matter are the method, meaning direct or indirect, and the horizon, meaning short, medium or long. Search results sometimes describe qualitative, quantitative, causal and time-series approaches, but those come from demand forecasting rather than from cash. For an operator the only two decisions are which method you are using and how far out you are looking, and both are covered on this page.
What tools can I use for cash flow forecasting?
A spreadsheet, for longer than most vendors suggest. We do not name products on this page, and we did not verify any pricing, ratings or free-tier limits, so nothing here recommends software or advises against it. The signal that a spreadsheet has been outgrown is not untidiness, it is two people producing different answers for the same cash position. That decision, with the staging by company size, sits on the wider guide to cash flow management rather than here.
Can I build a cash flow forecast in Excel?
Yes, and most of the businesses that do this well never use anything else. The structure is one column per week across thirteen weeks, one row per cash line, an opening balance at the top and a closing balance formula at the bottom that feeds the next column. Keep confirmed items and assumptions in visually separate blocks so nobody reads a guess as a fact. Add a variance column next to the first week so last week's forecast and last week's actual sit side by side, because that comparison is the part most sheets are missing.
Do I forecast the invoice date or the date the customer actually pays?
The date they actually pay, based on how that specific customer has behaved. Terms are a statement of intent and the aged receivables report is a record of reality, and where the two differ the record wins. A customer on 30-day terms who has paid on day 41 for six consecutive invoices belongs in week six, not week four. Forecasting the terms date rather than the observed date is the most common single reason a forecast is optimistic in the same direction every week.
Which week does payroll tax money actually leave the account?
That depends on which deposit schedule applies to you, and the schedule is set by federal rule rather than by your preference. IRS Publication 15 sets out two: monthly schedule depositors deposit by the 15th of the following month, while for semiweekly depositors liabilities incurred on Wednesday, Thursday and Friday are due the following Wednesday, and liabilities incurred on Saturday, Sunday, Monday and Tuesday are due the following Friday. Which one applies is determined by a lookback period, which the publication describes as the four quarters of the preceding year, July 1 through June 30. The practical consequence for a weekly sheet is that payroll and the tax on that payroll can land in different weeks. Confirm your own schedule with your accountant or payroll provider rather than with a web page.
How accurate should a cash flow forecast be?
Accurate enough to change a decision, which is a lower bar than most people assume. The first two weeks should be close, because they are built almost entirely from documents, and accuracy should visibly decay across the horizon because the later weeks contain assumptions. A forecast that is consistently wrong in the same direction has a broken assumption rather than bad luck, and that is fixable. No credible authority publishes an accuracy target for small business forecasting, and we are not going to invent one.
What is the difference between a cash flow forecast and a budget?
A budget answers whether the year works. A forecast answers whether next Tuesday works. The budget is built from targets and plans, usually in monthly columns across twelve months, and it is revised occasionally. The forecast is built from invoices, bills and contracts, in weekly rows across a quarter, and it is extended every week. They are complementary and they are not substitutes, and using a budget as a solvency instrument is how a business runs out of money in a year it hit its numbers.
Is a cash flow forecast the same as a statement of cash flows?
No, and the difference is direction. A statement of cash flows is a reporting document that explains where cash went during a period that has already finished, and under ASC 230 it classifies movements as operating, investing or financing. A cash flow forecast points the other way, at weeks that have not happened yet, and its output is a dated balance rather than a period result. They share vocabulary and answer opposite questions. The classification side of it is covered in the wider guide to cash flow management.
How often should I update the forecast?
Weekly, on the same day, as a fixed thirty-minute block. Reconcile the bank, enter what actually happened last week, read the difference against what you predicted, then extend the horizon by one week so it never shortens. The extension is what makes it rolling, and the comparison is what makes it improve. A forecast rebuilt monthly is a document; a forecast extended weekly is a practice, and only the second one catches anything.
What do I do in week one if the forecast says week six is short?
You have five weeks, which is the entire point of building the thing. In the worked example on this page, three moves close a $9,100 trough: collecting two overdue invoices, taking deposits on scheduled work, and moving a materials payment out on longer supplier terms. Only the first brings in money the business did not already have. The other two shift the timing, which is enough when the problem is timing. If the gap does not close on the sheet, the honest next step is a certified public accountant or an SBA resource partner rather than another tactic.
Where can I get a free cash flow forecast template?
Not from us, and it is worth saying why rather than sending you to a form. Template demand on this topic is large, with four Excel-template variants measuring 720, 720, 480 and 480 US searches a month on 18 August 2026, and almost every page serving it puts the file behind an email capture. This publication does not gate downloads. The eight-week table on this page is the template: nine input rows, one column per week, an opening balance at the top and a closing balance that carries forward, which is reproducible in a spreadsheet in about ten minutes.
Should a forecast include owner draws?
Yes, on their dates, at the amount actually taken. Owner draws leave the bank account exactly like any other payment, and leaving them out is one of the most common reasons a forecast shows a surplus while the account shows a shortfall. They belong in the outflow rows even though they are not a business cost, because the forecast is a record of money movement rather than of profitability. In the worked example they appear twice, at $3,000 each, in weeks one and five.