The most useful short-term accounts receivable forecast formula is invoice-level: validate each open balance, assign only supported amounts to plausible collection weeks, then sum the lower and upper amounts by week. Do not calculate expected cash by multiplying the entire aging balance by one generic percentage. That hides disputes, credits, payment promises, and timing exceptions that determine what can actually arrive.

Start with a controlled invoice population

Use one dated detail export. For each invoice record the customer, invoice identifier, open balance, due date, known dispute or credit, latest follow-up evidence, owner, and timing assumption. The reported open balance should already reflect posted payments and credits; do not subtract them a second time.

The invoice-level control total is:

`validated open AR = sum of validated invoice open balances plus separately identified non-invoice AR rows`

Reconcile that amount before forecasting. The four-week AR forecast control total explains how to keep forecast scope and accounting scope separate. If the source does not reconcile, retain the difference as an explicit exception rather than forcing it into a collection week.

Calculate a weekly range

For each invoice assign three fields: earliest supported week, latest supported week, and supported amount. A dated payment commitment may justify a narrower window than an overdue status alone. A dispute, missing contact, or unresolved application may justify no collection week yet.

For week `w`, use:

`lower bound(w) = sum of supported amounts whose conservative week is w`

`upper bound(w) = lower bound(w) + sum of additional amounts with credible but unresolved timing in w`

The four-week collections range is the cumulative lower and upper total across weeks one through four. Each invoice amount should appear at most once in each scenario. An amount outside the horizon remains in open AR and the action queue but not in the four-week expected-cash total.

This structure is compatible with the documented forecasting approaches described by Oracle and Microsoft, while keeping the short-term decision at invoice level. See the broader accounts receivable forecasting methods comparison before choosing a method for a different horizon.

Keep judgment outside the arithmetic

The summation should be deterministic. The judgment lies in whether evidence supports an amount and week. Record a reason code such as confirmed promise, normal terms with recent acknowledgment, disputed, payment pending, or no current evidence. Apply collection confidence labels consistently, but do not convert a label into false precision.

Review three tie-outs every cycle:

  • source rows to validated open AR;
  • invoice scenario allocations to each weekly subtotal; and
  • four-week subtotals to the reported lower and upper range.

Show the cash-receipts bridge explicitly

The weekly formula should make the relationship between open AR and forecast cash receipts visible:

`week cash receipts = sum of invoice amounts assigned to that week under the selected evidence scenario`

Then reconcile the full source population:

`open AR = four-week scenario amounts + later amounts + unassigned amounts + documented scope differences`

Keep cash already observed before the snapshot cutoff out of the future-week sum. Keep a deposit that has not yet been linked to a specific invoice in a reconciliation exception rather than subtracting it from whichever invoice looks most likely. The QuickBooks deposit-link guide shows the evidence needed before changing the invoice balance used by the formula.

This formula creates an auditable forecast rather than a guaranteed prediction. AR Pulse analyzes an authorized QuickBooks receivables export and turns the evidence into a ranked action list and four-week collections range. It does not contact customers or promise a payment outcome. View the example report to see the output boundary.