Cash flow forecasting is the practice of writing down, ahead of time, when money will arrive in your account and when it will leave. That’s it. You start with the cash you have, add what you expect to collect, subtract what you expect to pay, and read the closing balance for each week ahead.
The output is one number per period: how much cash you’ll have left. The value is entirely in seeing that number before the week arrives rather than after.
This matters more when your money moves on two different clocks. If you deliver your own orders, you pay drivers and buy fuel this week, and the wholesale customer who received those orders pays on day 30 or day 45. A profitable month can still empty the account mid-month. A forecast is how you find that week in advance, and once you’ve found it, the guide to managing debt in a delivery-heavy business covers the decisions you make about it, from drawing on a line of credit to restructuring a payment before it’s due.
This post is about the forecast itself: what goes in it, how far out to run it, how to build it, and how to keep it honest.
The Bottom Line
- A cash flow forecast tracks cash movement and timing. A profit and loss statement tracks earnings. A profitable month and an empty account are entirely compatible, which is why you need both.
- Use a rolling 13-week forecast in weeks, not months. Monthly buckets average away the exact squeeze you’re trying to find.
- Build it from payment dates, not invoice dates. What matters is when the money lands, and 47% of small businesses report invoices overdue by more than 30 days (Intuit QuickBooks, survey of 2,487 US small businesses, January 2025).
- Update it weekly by replacing estimates with actuals and rolling the horizon forward one week. An unmaintained forecast is worse than none, because you’ll trust it.
- The forecast’s job is lead time. Half of small businesses hold a cash buffer covering about 27 days of outflows (JPMorgan Chase Institute, 597,000 firms, February–October 2015), so a few weeks of warning is often the whole difference.
Save 80% of delivery management time
We handle everything:
- Dedicated operations manager
- Real-time tracking dashboard
- Automated customer notifications
- Urgent issue resolution
What a cash flow forecast is, and how it differs from a budget
A cash flow forecast answers one question: what will my bank balance be at the end of each week from here?
It’s easy to confuse with two documents that look similar and do different jobs.
- A profit and loss statement reports what you earned and spent over a period that has already happened, on an accrual basis, with revenue counted when invoiced rather than when paid. It can show a strong month while your account is nearly empty.
- A budget is a plan for what you intend to spend. It’s a target you’re measured against. A forecast is a prediction of what will actually happen, and it changes whenever reality does.
- A cash flow forecast is forward-looking, cash-only, and organized by date of payment. Nothing non-cash belongs in it. Depreciation doesn’t appear. An invoice you’ve sent doesn’t appear until the week you expect to be paid.
The practical consequence: an invoice raised on the 3rd for $8,000 with net-30 terms is revenue in March and cash in April. On the forecast it sits in the April week you actually expect the transfer, which for a slow-paying account might be the first week of May.
How far ahead should a cash flow forecast run?
Match the horizon to the decision you’re making. Most small businesses need two, not one.
| Horizon | Granularity | What it answers |
|---|---|---|
| 4 to 13 weeks | Weekly | Can I cover payroll, fuel, and this month’s loan payment? Which specific week gets tight? |
| 6 to 12 months | Monthly | Can I afford a second van, a lease, a hire? How deep is the slow season? |
| 1 to 3 years | Quarterly | Financing conversations and expansion planning. Directional only. |
The short-term weekly forecast is the one that prevents problems. The longer-term monthly view is the one that informs commitments. If you only build one, build the weekly.
Resist the temptation to run a weekly forecast out twelve months. Precision you don’t have isn’t useful, and maintaining it will be the reason you stop maintaining it.
What goes in the inflow and outflow lines
List the categories that are material for your business, not every category that exists. Eight to fifteen lines is normal. Anything smaller than roughly 1% of monthly cash movement can be grouped into “other”.
Cash in:
- Customer payments from invoices, grouped by expected payment week
- Point-of-sale and card receipts, which arrive on a settlement delay of one to three days
- Deposits on events, custom work, and large wholesale orders
- Loan or line-of-credit draws
- Owner contributions, grants, tax refunds
Cash out:
- Payroll and payroll taxes, on their actual dates
- Driver pay and delivery costs
- Fuel, tolls, parking, vehicle maintenance
- Inventory and ingredient purchases, on supplier terms
- Rent, utilities, insurance, software
- Loan and lease payments, including vehicle and equipment notes
- Sales tax and estimated tax payments
- Card balance payments
Two lines get missed most often. Quarterly and annual payments (insurance renewals, tax instalments, equipment servicing) are invisible in a monthly average and brutal in the week they hit; put each one on its actual date. And payment processing fees, which for a card-heavy business quietly run 2–3% of receipts.
How to build a cash flow forecast in six steps
Allow ninety minutes the first time. Twenty minutes a week after that.
- Set the opening balance. Today’s actual bank balance, from the bank, not from your books. If you hold several accounts, use the total you can actually reach today.
- Lay out your columns. One per week, thirteen of them, starting with the current week. Label them by date rather than week number so you can see the renewals and tax dates land.
- Fill in known outflows first. These are the reliable numbers: payroll, rent, loan payments, lease payments, insurance. They’re contractual, so they go in at full value on exact dates. Doing outflows first keeps the exercise honest, because inflows are where optimism lives.
- Add expected inflows by payment date. Go through your unpaid invoice list and place each one in the week you actually expect the money. Use each customer’s payment history, not their stated terms. A wholesale account that has paid on day 45 for two years pays on day 45.
- Add variable costs using recent averages. Fuel, driver pay, and inventory usually scale with delivery volume. Take the last eight weeks, work out a cost per order or cost per route, and apply it to your expected volume for each week ahead.
- Calculate the closing balance for each week, and carry it forward. Opening balance plus inflows minus outflows equals closing balance, which becomes the next week’s opening balance. That chain is the forecast.
A four-week extract looks like this:
| Week 1 | Week 2 | Week 3 | Week 4 | |
|---|---|---|---|---|
| Opening balance | 14,200 | 11,650 | 6,900 | 9,400 |
| Customer payments | 9,800 | 6,200 | 13,900 | 8,100 |
| Card settlements | 3,100 | 3,400 | 2,900 | 3,200 |
| Total in | 12,900 | 9,600 | 16,800 | 11,300 |
| Payroll and driver pay | 8,900 | 8,900 | 9,600 | 8,900 |
| Fuel and vehicle | 1,450 | 1,500 | 1,700 | 1,450 |
| Inventory | 4,100 | 3,950 | 4,800 | 4,100 |
| Rent and utilities | — | — | 3,200 | — |
| Loan and lease payments | 1,000 | 1,000 | 1,000 | 1,000 |
| Insurance renewal | — | — | 4,000 | — |
| Total out | 15,450 | 15,350 | 24,300 | 15,450 |
| Closing balance | 11,650 | 6,900 | 9,400 | 5,250 |
Illustrative figures. The point is what the layout reveals: week 3 looks alarming and recovers, because rent and an insurance renewal land in the same week as a large collection. Week 4 is the real problem. Nothing unusual happens, and the balance still drops to 5,250 on a 15,000 weekly outflow. That’s a third of a week’s cushion, visible three weeks early.
Why a 13-week rolling forecast became the standard
Thirteen weeks is one quarter, and a quarter is long enough to see a seasonal turn coming while staying short enough that your estimates mean something. It’s the horizon lenders, accountants, and turnaround specialists ask for.
“Rolling” is the more important half of the phrase. Each week you do two things: replace last week’s estimates with what actually happened, and add a new thirteenth week at the far end. The horizon stays thirteen weeks wide forever, so you never run out of runway and never rebuild from scratch.
Weekly granularity is what does the work. A monthly forecast that shows $40,000 in and $38,000 out looks fine and can still contain a week where payroll and a tax payment land before the month’s largest collection. Averaging hides precisely the event you’re forecasting to avoid.
How to forecast delivery costs that change week to week
Fixed costs are easy. The delivery line is the one that moves, and getting it approximately right is most of forecast accuracy for a business that ships its own orders.
Build it from a unit rate rather than a monthly figure:
- Work out your cost per delivery or per route. Total driver pay, fuel, tolls, and vehicle costs for the last eight weeks, divided by orders delivered or routes run. One number you can multiply.
- Forecast volume by week, not by month. Your own order history from last year is the best predictor you have. A florist’s February, a caterer’s December, and a bakery’s holiday weeks are not the average.
- Add the peak premium separately. Peak weeks often cost more per order, not just more in total: overtime, a rented vehicle, a same-day courier for the order that didn’t fit. If you covered peak with per-delivery capacity rather than a second van, that cost is variable and predictable, which is the easier shape to forecast.
- Budget a redelivery rate. Failed deliveries cost fuel and labor twice and push the payment later. If yours run 2%, forecast 2% rather than zero.
Fuel deserves its own line rather than being folded into vehicle costs, because it’s the one input whose price moves independently of anything you do. If your routes are optimized for distance, that line is both smaller and steadier, which shows up directly in forecast accuracy.
How to keep a cash flow forecast accurate
Accuracy comes from one habit: comparing what you predicted to what happened, every week, and asking why they differ.
Keep last week’s forecast next to the actual result and look at the variance line by line. You’re not chasing perfection. You’re looking for patterns.
- A customer who consistently pays later than you forecast: change the assumption for that customer permanently. Most forecast error is a handful of accounts you keep giving the benefit of the doubt.
- A cost category that’s always higher than forecast: usually fuel or inventory, usually because you used an old unit rate.
- Recurring payments you keep forgetting: once identified, they belong on the calendar, not in your memory.
A forecast within 5–10% on closing balance across a thirteen-week horizon is doing its job. Chasing exactness is how the habit dies; ten accurate minutes a week beats an elaborate model you abandon in March.
Two more things protect it. Forecast conservatively on the way in and generously on the way out. If a payment could arrive in week 2 or week 4, put it in week 4. And keep one version. Owners who maintain an optimistic forecast and a pessimistic one end up trusting neither.
What to do when the forecast shows a shortfall
The point of finding week 4 early is that you have four weeks of options instead of a decision made on the day.
In roughly the order worth trying:
- Pull cash forward. Invoice immediately on delivery rather than weekly, chase the two largest overdue accounts by phone, offer a small discount for payment in ten days.
- Push non-critical cash out. Ask a supplier for an extra fifteen days, delay a purchase by a week, move a discretionary payment past the tight week.
- Move a due date. Lenders will often shift a payment date to match when your receivables actually arrive, and it costs nothing to ask.
- Draw on the line of credit deliberately. A planned draw with a repayment week identified is a tool. An unplanned overdraft is a fee.
- Restructure the payment. If the shortfall is structural rather than a one-week timing issue, that’s a conversation with the lender, and it goes much better before a payment is missed than after.
The last two are debt decisions rather than forecasting ones, and the companion guide on managing debt covers how to size payments against seasonal cash and when refinancing is worth doing.
Worth noting how common the underlying situation is: in the Federal Reserve’s 2025 Small Business Credit Survey, meeting operating expenses was the most common reason firms sought financing, cited by 56% of applicants (Fed Small Business, 6,525 responses, fielded September 3–November 14, 2025). A forecast doesn’t remove the need to borrow. It changes borrowing from a reaction into a choice, which is usually the difference between a line of credit and an advance against daily receipts.
Common cash flow forecasting mistakes
- Using invoice dates instead of expected payment dates: the single most common error, and it makes every week look better than it is.
- Forecasting in months: averages away the shortfall you’re looking for.
- Leaving out quarterly and annual payments: insurance, tax instalments, and equipment servicing wreck a week that otherwise looked fine.
- Building it once: a forecast from six weeks ago is a historical document.
- Forecasting revenue instead of cash: revenue you haven’t collected pays nobody.
- Ignoring your own seasonality: last year’s weekly order history is sitting in your system and is a better predictor than any assumption you’ll invent.
- Forgetting card settlement delays and processing fees: small individually, material over thirteen weeks.
Frequently asked questions
What is the difference between a cash flow forecast and a cash flow statement?
A cash flow statement is a historical accounting report covering cash movement in a period that has ended, usually split into operating, investing, and financing activities. A cash flow forecast is forward-looking and built around your own payment dates. The statement tells you what happened; the forecast tells you what’s about to.
How far out should a small business forecast cash flow?
Thirteen weeks in weekly detail, updated every week, plus a rougher twelve-month monthly view for decisions like hiring or financing a vehicle. The thirteen-week model prevents cash emergencies. The twelve-month view informs commitments. Anything beyond a year is directional.
Do I need software to build a cash flow forecast?
No. A spreadsheet with dated weekly columns is enough, and it’s what most small businesses use. Software helps by pulling your invoice and bill data automatically, which mainly saves the ten to twenty minutes of weekly data entry and reduces typing errors. The accuracy still comes from your assumptions about when customers pay, and no tool knows that better than you do.
How accurate should a cash flow forecast be?
Within 5–10% on the closing balance is good for a thirteen-week horizon, with the nearest two or three weeks noticeably tighter than the far end. If you’re consistently off by more, the cause is almost always customer payment timing rather than your cost estimates. Track the variance weekly and fix the assumptions that keep being wrong.
What’s the most useful single number in a cash flow forecast?
The lowest closing balance across the horizon, and the week it falls in. Everything else is supporting detail. If that number is below one week of outflows, you have a problem to solve, and you’ve found it with weeks to spare.
Start with thirteen columns
Cash flow forecasting isn’t an accounting exercise, and it doesn’t take an accountant. It’s a spreadsheet, your unpaid invoice list, your loan payment dates, and an honest view of when each customer actually pays.
Open a blank sheet. Put today’s bank balance in the first cell. Add thirteen weekly columns, fill in the contractual outflows, then place each unpaid invoice in the week you really expect the money. The lowest closing balance you see is the number that matters, and knowing it three weeks early is what the exercise buys you.
Then update it next week, and the week after. The habit is worth more than the model.
