Most 13-week cash forecasts do not govern decisions. They are produced, distributed, filed, and not used until the next one is produced.

This is not because the format is wrong. The format is mostly fine. It is because the operating context the forecast is supposed to drive has not been built around it. A 13-week forecast that governs decisions has six structural and behavioral attributes that most do not.

What “governing decisions” actually means

A cash forecast governs decisions when:

  • The CEO and CFO open it weekly, not monthly.
  • The sponsor and lender accept it as the working liquidity view, not a formality.
  • AP release timing, capex pacing, hiring decisions, and customer collections are sequenced against it.
  • Variance to forecast triggers a defined response — not a discussion.

A forecast that does none of those things is a finance artifact. It exists. It does not run anything.

Why most 13-week forecasts fail

The default version fails for predictable reasons.

It is built once and updated mechanically. The first week’s forecast was built carefully. By week six, it is being rolled forward by inertia. Driver assumptions are not re-examined. The forecast becomes a smoothed projection, which by definition cannot govern decisions because it does not surface anything.

It is not driver-based. Cash receipts are forecast as a single line. AR aging, customer concentration, and payment behavior are not inputs. When receipts come in low one week, no one can explain why or whether it will recur, because the forecast does not connect to operating drivers.

It is owned by finance alone. The forecast lives in the FP&A team’s spreadsheet. AP and AR don’t see it. Operations don’t see it. Customer-facing teams don’t see it. So no one outside finance can act on it, and finance cannot act unilaterally.

It is the same length and shape as last week’s. No movement on the rolling 14-week edge. No changes in the assumption sheet. No annotations on what changed. Sponsors and lenders read this as performative — produced because it must be, not because it informs.

There is no consequence to being wrong. When variance happens, no one is asked to explain it. No process re-baselines. The forecast tolerates inaccuracy because no decision is keyed to it.

Structural requirements

A forecast that governs decisions has the following structural attributes.

Driver-based receipts. Customer receipts forecast at the customer or customer-cohort level, with payment behavior assumptions surfaced and named. AR aging buckets visible. Customer concentration risks called out. When a top-five customer slips a payment, the forecast moves and the reason is in the assumption sheet.

Disciplined AP timing. Vendor payments scheduled by category and term, with explicit decision points on which categories release and which hold each week. AP is not a projection — it is a managed line.

Named operating inputs. Payroll cycles, sales tax remittances, debt service, lease payments, and capex commitments are treated as scheduled outflows with explicit timing, not estimates.

Covenant compliance walkforward. Where applicable, the forecast carries the covenant calculation forward week by week. The lender does not have to ask. The number is in the package.

Variance tracking. Each week’s actuals are reported against last week’s forecast for that week. Variance is decomposed into receipts, AP timing, and other. Rolling 4-week and 13-week variance trends are visible. Persistent variance triggers re-baselining of the underlying driver, not a softer assumption.

Scenario layer. A base case, a downside case (typical drivers: customer slip, margin compression, supply chain disruption, financing delay), and the implied actions in each. Not three full models — three deltas off the base, with the operating responses already named.

Operating cadence

Structure alone does not make a forecast govern decisions. The cadence does.

A forecast that governs decisions runs on a defined weekly cycle. A representative cadence:

  • Tuesday. Treasury and AR close out the prior week’s actuals; AP team confirms the upcoming week’s release schedule.
  • Wednesday. FP&A updates the forecast, reconciles last week’s variance, refreshes assumptions, distributes the package.
  • Thursday. CFO and CEO review with named operating leads (commercial, operations). Decisions are made on AP holds, capex pacing, and any required sponsor or lender communications. Decisions are recorded.
  • Friday. AP releases against the agreed schedule. Customer collections team receives the priority list.
  • The following Tuesday. Variance is reported back. Where forecast missed, the driver is identified and the assumption is reset.

The package is short — under five pages. The meeting is short — under thirty minutes. Both happen every week without exception. When the cadence is broken once, the forecast loses authority. When it is broken twice, the forecast is dead and a new one will need to be stood up to recover the discipline.

A forecast built this way has external value beyond internal management.

Sponsors read it as a leading indicator. A forecast with named driver assumptions, weekly variance tracking, and visible decision-making is a stronger credibility signal than any monthly board package narrative. It tells the sponsor that management can see the business clearly and act on what it sees.

Lenders use it for covenant tracking and downside scenario testing. A lender provided a covenant compliance walkforward inside the cash package every week does not need to call. A lender who has to ask is already losing confidence.

In a stressed credit, the forecast becomes the working document of the relationship. Lenders move from monitoring to closer monitoring to workout in part based on whether they trust management’s liquidity view. A forecast that has governed decisions for six months before a stress event carries more weight in that conversation than one assembled in response to it. The forecast is not a document the company produces during stress. It is a document the company produces against the possibility of stress.

Common failure modes

Three failure modes show up repeatedly.

Forecast inflation. Receipts forecast aggressively to make the picture look acceptable. The forecast is then chronically optimistic, variance compounds, and credibility decays. The fix is mechanical: forecast receipts at the customer level using observed payment behavior, not management-preferred outcomes.

Capex held off the forecast. Capex commitments treated as flexible when they are contractually binding. When the bill comes due, the forecast misses by the capex amount and no one anticipated it. The fix: capex commitments enter the forecast at the moment the PO is signed, not the moment the invoice arrives.

Inventory ignored. Inventory build or release decisions made by operations without finance line of sight to the cash impact. The forecast misses; the explanation arrives a week late. The fix: a weekly inventory line in the forecast, reconciled to operations’ build plan, with the cash impact named.

The fix in each case is the same: name the driver, assign the owner, surface the assumption, track the variance.

A 13-week cash forecast that does these things stops being a finance exercise. It becomes the instrument the company runs on for the period it matters most.