Define Week 1 as a documented, closed date interval that begins immediately after the forecast snapshot cutoff. Then define Weeks 2 through 4 as the next three non-overlapping intervals of the same type. Do not let “this week,” “next week,” and the report as-of date shift meaning between reviewers.
Microsoft's cash-flow forecasting documentation distinguishes several dates that can affect cash timing, including invoice, payment-term, due-date, and payment-delay inputs. Intuit's A/R aging guidance likewise makes the report date and transaction detail part of the aging view. A short-term collections forecast needs one additional control: a precise mapping from calendar dates into forecast weeks.
Write the boundary rule on every snapshot
For each forecast version, preserve:
| Control | Example | Purpose | | --- | --- | --- | | Snapshot cutoff | Friday, 5:00 p.m. local time | Separates known facts from later events | | Week 1 start | Saturday | Opens the first forecast interval | | Week 1 end | Following Friday | Closes the interval without overlap | | Time zone | Company reporting time zone | Prevents midnight ambiguity | | Calendar rule | Seven-day weeks or named business weeks | Keeps later snapshots comparable |
The example dates are not a universal prescription. A Monday-through-Sunday operating week can work just as well. The control is consistency: one date belongs to one interval, and the same rule is used when actual receipts are compared with the locked forecast.
Treat the snapshot date separately from Week 1
The snapshot records what was known at the cutoff. A receipt observed before the cutoff belongs in actual or recorded cash, not in a future forecast week. A receipt first observed after the cutoff may be a Week 1 outcome even if the customer's payment instruction was dated earlier.
This separation prevents two common errors: counting the same cash in the opening state and Week 1, or moving a late-discovered receipt backward into the locked forecast. The cash-receipt cutoff guide explains the evidence check for recorded receipts.
Assign dates before assigning confidence
Map a supported expected date into the calendar boundary first. Then label evidence quality separately. A high-confidence invoice and a low-confidence invoice can share a week, but they should not be collapsed into the same certainty claim.
If the evidence provides only a range, keep that range visible. Do not force the invoice into Week 1 because the earliest possible date lands there. Use an unassigned or cross-week state until the review rule supports one treatment.
Roll forward without rewriting history
When the next snapshot is created, close the prior Week 1, record observed outcomes, and shift the remaining horizon under a new version. Preserve the old boundaries and assumptions. The rolling-versus-frozen forecast guide shows why the operating view and the measurement baseline should remain separate.
The accounts receivable forecast template can carry the cutoff, week-start, week-end, evidence, owner, and outcome fields together. Before release, verify that every expected receipt falls in exactly one weekly, later, or unassigned state.
AR Pulse analyzes an authorized QuickBooks receivables export to create a ranked cash-action list and a four-week collections range. It does not require QuickBooks credentials, change books, or promise payment outcomes. View the one-page example to see how source facts and timing assumptions stay separate.