Manufacturing orders: test backlog aging and conversion
Manufacturing backlog increases when new orders exceed the orders removed through fulfillment or other adjustments. That identity does not tell you whether demand strengthened or production slowed. Combine a stock-flow reconciliation with aging buckets so the same ending balance can be interpreted through its composition.
Anchor the identity to the data system
The US Census unfilled-orders survey describes new orders as shipments plus the change in unfilled orders. This is a statistical construction within that survey’s coverage and benchmarking process. A particular company’s bookings, shipment and recognized-revenue definitions may differ, so reproduce its own reconciliation before using the identity.
Inspect whether orders are net of cancellations and whether acquisitions, price changes or foreign exchange alter the backlog. A backlog stock at month-end should not be compared with year-to-date shipments as though both represented one month’s activity. Seasonally adjusted industry data also differ from an issuer’s reported results.
Hypothetical backlog with aging
Assume an invented manufacturer begins a quarter with backlog of 400 million, receives 180 of net new orders and ships 150. With no other changes, ending backlog is 430. Its book-to-bill ratio using these definitions is 180 / 150 = 1.2.
Now split the ending stock by the promised delivery date. Of the 430, 260 is not yet due, 90 is overdue by up to ninety days and 80 is overdue longer than ninety days. A year earlier, the comparable long-overdue bucket was only 20. The total grew, but the composition raises a fulfillment question.
| Aging worksheet | Ending balance | Next evidence |
|---|---|---|
| Not yet due | 260 | Scheduled production slots |
| Overdue up to ninety days | 90 | Parts availability and rescheduling |
| Overdue more than ninety days | 80 | Customer acceptance of revised terms |
| Total backlog | 430 | Reconcile with opening stock and flows |
The aging buckets are based on promised delivery, not simply the original order date. A legitimate eighteen-month contract should not be labeled delinquent after three months. Conversely, repeatedly resetting a promised date can make a troubled order appear current. Preserve the original commitment and the latest agreed schedule separately.
Distinguish bottlenecks from weak orders
For late orders, record the limiting component, work-in-process stage, customer changes and cancellation rights. A component shortage can delay otherwise viable demand. A customer repeatedly postponing delivery may instead indicate lower urgency or excess capacity. The same order value cannot distinguish those explanations.
Compare factory output with shipments and finished-goods inventory. If completed goods accumulate while shipments stall, the issue differs from an upstream component bottleneck. If order values rise because prices increased, physical unit demand may be flat. Use unit data where disclosed and leave missing detail explicit.
Define the follow-through test
The next-quarter check should ask how much of the long-overdue 80 shipped, was canceled, was renegotiated or remains unresolved. Do not merely inspect whether total backlog rose again. A stronger new-order quarter can mask deterioration in old orders unless their outcomes remain visible.
Public reporting rarely provides a complete aging ledger. In that case, use disclosed delivery delays and inventory patterns as partial evidence rather than inventing exact bucket amounts. The hypothetical table is a reusable research structure, not a claim that the survey or issuer publishes these fields. It identifies an operating question and does not forecast manufacturing revenue.
FAQ
Does book-to-bill above one prove growth?
It proves that defined orders exceeded defined billings or shipments during that period. Cancellations, price and fulfillment timing still determine its interpretation.
Should all old orders be discounted identically?
No. Compare their contractual delivery dates, customer status and production stage. Age alone is insufficient to estimate an economic loss.