A cash flow forecast projects receipts and payments forward to a closing balance per period, so a shortfall shows up while it can still be fixed. Pick a horizon that matches how much certainty you have, carry a costed downside alongside the base case, and give one person the job of keeping it current.
What goes into a cash flow forecast
Four inputs, and the discipline is in the timing rather than the amounts. Receipts from customers, dated by when you expect the money to land rather than when you invoiced. Payments to suppliers, dated by terms. Payroll and its on-costs, which are the most predictable line on the sheet and the least forgiving. Then everything with a statutory due date, chiefly your BAS obligations and superannuation, which arrive whether the quarter went well or not.
A forecast built from an aged receivables report and supplier terms will be roughly right immediately. A forecast built from the budget divided by twelve will be wrong in a specific and dangerous way, because it smooths exactly the lumpiness that causes the problem. The general framing in business.gov.au's cash flow guidance is a reasonable starting structure.
The more complete version links the forecast to the profit and loss and the balance sheet so the three move together. Kaycee Singh described teaching herself three-way forecasts from YouTube videos at a health startup, because cash burn was the absolute focus and accurate cash forecasting was critical.[1] That is a fair picture of how this capability arrives in most scale-ups: someone works out they need it, and builds it.
Choosing the horizon
There is no single right period, only a match between horizon and how much certainty you actually have. Michael Oraniuk described how during challenging times his forecasting horizon extended a month, then three months, and never further than about three, requiring a really close week-by-week focus.[2] That is not a failure of planning. It is a correct read of what was knowable.
| What it answers | When you need it | |
|---|---|---|
| Rolling 13 weeks, weekly | Can we make payroll, and which specific receipt is the one that has to land? | Whenever cash is tight, a raise is in progress, or receipts are concentrated in a few large customers. |
| 12 months, monthly | Does the plan hold for the year, and where are the seasonal troughs? | Standard board reporting. This is the horizon a budget can genuinely support. |
| 18 to 24 months, monthly | When does the money run out, and when do we need to raise? | Venture-backed companies between rounds, where the answer sets the fundraising calendar. |
| Scenario overlays | What breaks first, and what would we do about it? | Run alongside any horizon, and worth building before conditions turn rather than during. |
Most growing companies should run two at once: a short weekly view for operations and a longer monthly view for the board. They will disagree. That is fine, and the difference between them is usually the most informative number in the pack.
The line the forecast is measured against is your buffer. Here is how to size a business cash reserve so the threshold means something.
Scenarios, and running them before you need them
A single-line forecast is a prediction, and predictions are wrong. The useful version carries a downside you have actually costed. Rob Doyle described running forecast scenarios during COVID that included people being unable to transact property for six months, which prompted the business to secure additional debt facilities and safeguard operations, and reflected afterwards that they could have put the foot down harder.[3] The point is the timing: the facilities were arranged while the business still looked lendable.
Keep the scenario set small and specific. One downside on revenue, one on collection timing, one on a single large customer or contract, each expressed as a change to a driver rather than a percentage haircut on the total. And name the action each one triggers in advance. Sean O'Donoghue described the response to financial pressure as knuckling back down to fundamentals like cash flow and forward workload, then deciding what to cut back, defer, eliminate or stop for a period.[4] Those four verbs are the decision menu.
Reading the output
The number that matters is the projected closing balance in each period, and the first period where it falls below your threshold. The threshold should not be zero. Set it at the point where you would need to start making decisions you would rather not make, which for most businesses is one payroll cycle plus the next BAS.
Then treat the first breach date as a countdown rather than a fact. The forecast is worth running only if something changes when it turns red: chasing a receivable, deferring a hire, drawing a facility, or starting a raise. A forecast that produces no decisions is a report, and there are cheaper ways to produce those.
Who should own the forecast
Understanding the cash position in the first period of a new role is a critical focus for any incoming finance professional.[5]That is a fair marker for what the job is: not building the model once, but owning the number every week, and being the person who says out loud when the date has moved.
Founders can run a forecast, and many do it well. What they cannot do is keep it current alongside running the company, which is why the model tends to be accurate the month it was built and directionally wrong six weeks later. Fractional CFOs are a classic startup answer here, because businesses often need someone to get the house in order in the lead-up to a raise, focused specifically on visibility and runway.[6] That is a genuine fit for a defined period.
Beyond that period it becomes a hire. Story places two: the first finance hire at roughly 10 to 20 people, and the first CFO at 50-plus heads. If the board is asking for a rolling forecast and scenario commentary and nobody currently owns it, that is the signal, not the headcount.
On the timing question specifically, here is when to make your first CFO hire and what triggers it.
Common questions
What is a cash flow forecast?
A cash flow forecast projects money entering and leaving the bank account over a future period and ends in a projected closing balance for each period. It is a management tool, not the statutory cash flow statement, and the thing that separates a useful one from a useless one is dating rather than content. Date each receipt by when the money is expected to land, not by when it was invoiced. Get that wrong and the annual total still comes out right while the shortfall shows up in the wrong month, which is the one job the forecast had.
How far ahead should a cash flow forecast go?
As far as you have real certainty, and no further. The horizon is set by the decision you are making rather than by convention: chasing a receipt needs weeks, a board pack needs the year, and a fundraising calendar needs eighteen months or more. Running one long forecast and reading its twentieth month with the same confidence as its second is the common mistake. Later months are a shape rather than a number, and a runway date lifted out of one is how a board ends up planning against arithmetic nobody believes.
What is a three-way forecast?
A three-way forecast links the profit and loss, the balance sheet and the cash flow forecast so they move together, meaning a change in an assumption flows through all three consistently. That makes it more work to build than a standalone cash forecast and considerably harder to fool, because the balance sheet has to keep balancing. It is the version investors and lenders expect from a company past the earliest stage.
Who should own the cash flow forecast in a startup?
One named person whose job includes keeping it current, not just building it. The trigger for making that a hire is not headcount or revenue. It is the board asking for a rolling forecast with scenario commentary and nobody in the room being able to say whose job that is. Before that point a founder or a fractional CFO carries it perfectly well. After it, the forecast is being used to make decisions somebody has to answer for, and the person answering needs to be the person maintaining it.
References
- Kaycee Singh, who built a tech finance function from zero, on The CFO Track: she taught herself three-way forecasts using YouTube videos at Midnight Health, because startup cash burn was an absolute focus point and accurate cash forecasting was critical.
- Michael Oraniuk on The CFO Track: during challenging times his forecasting horizon typically extended only a month, then three months, but never further than about three, requiring a really, really close focus, week by week.
- Rob Doyle on The CFO Track: during COVID-19 his team ran forecast scenarios including the possibility that people could not transact property for six months, prompting them to secure additional debt facilities and safeguard operations, and he reflected that they could have put the foot down a bit harder.
- Sean O'Donoghue on The CFO Track: when facing financial challenges, knuckle back down to fundamentals like cash flow and forward workload, then make decisions on what to cut back, defer, eliminate, or stop for periods of time.
- Michael Oraniuk on The CFO Track, identifying understanding the cash flow situation in the initial period as a critical focus for finance professionals.
- How I explain it: they are a classic startup requirement, because businesses do not need someone full-time but have issues to sort out leading up to a raise, or need help with the raise process itself, focused on visibility and runway.