The 13-week cash forecast people actually use

Every business that has ever been close to running out of cash has, at some point afterwards, commissioned a thirteen-week cash forecast. Most of them sit in a spreadsheet that the controller updates dutifully every Monday morning and that nobody else opens. The forecast is not wrong, exactly. It is just decorative. It does not shape any decision the business actually takes during the week, and so it never quite earns the place it was supposed to occupy.

A thirteen-week forecast becomes useful only when it is integrated into the operating rhythm — when the conversations about which invoices to chase, which payments to defer, which hires to hold and which suppliers to renegotiate are anchored to it rather than running alongside it. This piece is about how to get to that state without launching a transformation programme to do it.

One owner, one cadence, one source of truth

The first failure mode of a thirteen-week forecast is shared ownership. Three people contribute, none is accountable, and inconsistencies between their inputs get reconciled in private rather than surfaced in public. The forecast then drifts and credibility erodes within six weeks.

Name a single owner. The owner is responsible for the forecast existing in a current state every Monday morning, with the previous week's actuals reconciled to the cash position and the next thirteen weeks rolled forward. Inputs come from named contributors with documented schedules — sales pipeline by Wednesday, payroll forecast by Thursday, capex commitments by Friday. The owner does not produce the inputs; they integrate them.

There is one source of truth. Not a folder of spreadsheets, not a working version and a published version. One file, one location, version-controlled, with the previous week's published version archived. The discipline of having a single artefact is what allows the conversation to stay anchored.

Build receipts and payments separately

Receipts and payments behave differently and need to be modelled differently. Receipts are driven by the order book, by historical conversion patterns, and by the collections process. Payments are driven by payroll, supplier commitments, capital commitments, and tax. Combining them into a single net cash flow line hides the levers that matter.

Receipts should decompose by channel where channels behave differently — direct sales, wholesale, retail, marketplace. Each channel has its own conversion curve from order to cash and its own collections behaviour. A consolidated receipts line will look stable while individual channels swing significantly underneath, and the swings will surface as variance you cannot explain.

Payments should decompose by category and by discretion. Payroll is non-discretionary and known in advance. Supplier payments are partially discretionary in timing if not in amount. Marketing spend is discretionary in both. Capex is discretionary in timing in most cases. The forecast should make the discretion visible, because that is exactly the lever the executive team will be asked to pull when cash gets tight.

Working capital is a separate line, not a residual

The most common modelling mistake we see is treating working capital as a plug — the line that absorbs whatever the receipts and payments lines do not explain. This guarantees that the working capital movement is unintelligible and that the model loses credibility the first time it gets challenged.

Model working capital explicitly. Days sales outstanding, days payable outstanding, inventory days where relevant. Show the movement in working capital as a separate line in the forecast, with assumptions visible. When DSO ticks up by three days, the cash impact should be visible on the page rather than buried in the residual.

Done this way, working capital becomes a managed lever rather than a source of variance. The first conversation we have with most clients about cash extension is not about cutting costs — it is about widening DPO and tightening DSO. Both can move materially within four weeks if the conversation is held professionally.

Three scenarios, each changing one cluster of assumptions

Scenarios are where most forecasts go wrong. The temptation is to produce many of them, with intricate combinations of assumptions, on the grounds that the future is uncertain and many outcomes are possible. The result is that no scenario is interpretable, because the differences between them are too multi-dimensional to discuss.

Hold yourself to three. A base case reflecting current run-rate and committed plans. A downside flexing one cluster of assumptions — typically conversion and average order value, or pipeline conversion to revenue, or supplier price movement. An upside flexing a different cluster — typically a planned channel expansion or a planned working capital improvement. Each scenario changes exactly one cluster, so the conversation in the executive office is always 'what changes if X' rather than 'why are these two numbers different'.

If the business genuinely needs a fourth scenario, it almost always means one of the existing scenarios is poorly framed. Reframe rather than add.

The Monday morning ritual

The forecast becomes part of the operating rhythm only when it has a fixed weekly review at which decisions are taken. The most effective form is a thirty-minute Monday morning slot with the CFO, the COO and the head of sales. The owner walks through the previous week's variance, the current cash position, and the forward thirteen-week view. The conversation focuses on three questions: what changed against last week's forecast, what decisions are pending, and what should we be worried about for the next four weeks.

Decisions taken in that meeting are logged with a named owner and a date. The next Monday's review begins by closing out the previous week's decisions. This rhythm is what turns the forecast from a passive artefact into the spine of cash management. Without it, the forecast is correct but unused.

Make the board see it weekly

The forecast should appear in front of the board weekly. Not in a board meeting — in a one-page email summary that the CFO sends every Monday afternoon. Cash position, forward thirteen weeks, the top three risks, the top three decisions taken or pending. Two minutes to read.

The discipline of producing that summary every week is what keeps the forecast honest. The CFO will not send a forecast they do not believe to a board on a Monday afternoon. The act of sending forces a level of rigour that monthly review does not.

It also gives the chair and the lead investor early visibility of strain. Cash problems escalate slowly until they don't. A weekly note means the board sees the trajectory rather than the cliff edge, and the conversations that need to happen — about a bridge round, about a cost cut, about a renegotiation — happen four weeks earlier than they otherwise would.

What you have when you finish

A thirteen-week cash forecast that is part of the operating rhythm is the difference between a business that runs out of cash gracefully and one that runs out of cash by surprise. The mechanics are not difficult. The discipline is. Most forecasts fail not because the model is wrong but because the rhythm around it is missing.

If you can answer three questions on any given Tuesday — what is the cash position, what is the forward thirteen-week view, and what decision are we taking this week to shape it — you have the forecast doing what it should. If you cannot, the forecast is still decorative.