Sizing the cut-off testing window for manufacturing clients with long dispatch lead times

A cut-off testing window has to be sized to how long it actually takes the client to get goods out the gate after invoicing, not to a default number of days. A retailer who ships same-day can use a narrow window, five to seven days either side of the period end, because an invoice dated close to year end is also dispatched close to year end. A manufacturer who invoices against a production schedule and dispatches two or three weeks later needs a window wide enough to catch invoices raised well before the period end that haven't actually gone out yet — and the reverse, dispatches just after year end against invoices raised weeks earlier. Get the width wrong and the window-based selection, however complete it is within its own boundaries, simply doesn't reach the transactions that matter.

Why manufacturing breaks a one-size window

In a make-to-order or make-to-stock-then-allocate business, invoicing is often tied to a production or allocation milestone, not to the goods physically leaving the factory. A unit can be invoiced the day it comes off the line and sit in the finished goods yard for ten, fifteen, twenty days waiting for a truck, an export document, or a customer's own readiness to receive. That gap between invoice date and dispatch date is exactly the gap cut-off testing is meant to probe — and a narrow window simply never sees the invoices sitting furthest out in that gap.

• Heavy engineering, machinery, and capital goods: dispatch frequently trails invoicing by two to four weeks, sometimes longer for export consignments awaiting documentation
• Textile and garment manufacturing on long production runs: batches invoiced together but dispatched in tranches over the following fortnight
• Pharma and chemicals with batch release testing: invoice can precede the quality-release date that actually gates dispatch
• Contract manufacturing with named-carrier pickup schedules: dispatch date is set by the carrier's route, not the client's warehouse

A practical way to size the window

1. 01Pull the average invoice-to-dispatch gap from last year's dataFor a sample of transactions from the prior period, compare the invoice date on the sales register to the dispatch date on the corresponding delivery challan or e-way bill. The spread, not just the average, tells you how wide the tail is.
2. 02Set the window to cover the tail, not the medianIf most invoices dispatch within five days but a meaningful minority take eighteen, a five-day window misses exactly the transactions most likely to straddle a period end. Size to the tail.
3. 03Widen further for the specific monthProduction schedules often bunch around month end regardless of normal lead time — a year-end push to ship everything invoiced in March before April 1 is common and argues for widening the window for that one period even if the rest of the year runs narrower.
4. 04Document the basis for the width chosenA reviewer should be able to see why fourteen days was chosen over seven, tied to the actual dispatch data, not a round number picked because it sounds reasonable.

What the wider window does and does not do

Widening the window increases the number of transactions flagged for tracing — that is the point, not a cost to minimise away. A fourteen-day window on a manufacturing client with genuine long lead times will surface more invoices than a seven-day window would, and that is the window doing its job of reaching the part of the population where a genuine cut-off error is most likely to sit. What the window does not do, at any width, is look at a dispatch note itself. Lekhya's revenue and cut-off testing flags every transaction within the configured window across the full population; it does not hold a dispatch-date field and does not compare invoice date to dispatch date on its own. Someone still pulls the delivery challan or e-way bill for each flagged transaction and makes the actual cut-off call.