The component list behind every manufactured item — and the single record that most often explains why product costing is wrong.
A bill of materials is the structured list of components, sub-assemblies and quantities required to make one unit of a finished item. In an ERP it drives material requirements planning, work order component issue and the material element of product cost. A BOM that is 95% accurate produces costing that is confidently wrong.
BOM accuracy is the quiet determinant of whether a manufacturing ERP tells the truth. Planning explodes demand through the BOM to decide what to buy; work orders issue components according to it; and standard cost is built from it. An out-of-date BOM therefore produces wrong purchase suggestions, wrong issues and wrong margins simultaneously — and nobody notices until year-end stock valuation.
Lists only the immediate components of one parent item. Simple to maintain and the usual building block.
The full indented structure down through sub-assemblies to raw materials. Useful for cost roll-up and for understanding lead time.
A sub-assembly that is never stocked — planning blows straight through it to its components. Common for kits assembled on the line.
A component line can carry an expected scrap percentage so that planning and costing reflect reality rather than theory.
A BOM line can start and stop on dates, which is how an engineering change is introduced without rewriting history.
A fabricator builds a steel enclosure. The BOM lists one sheet per unit, and for two years nothing seems wrong — the stores issue a little extra and the discrepancy is absorbed in the monthly stock adjustment.
In reality the cutting process loses about four percent to offcuts. On a thousand units a year that is forty extra sheets consumed and never planned. Purchasing buys to the BOM, so the shortfall appears as emergency orders at premium prices, roughly four a year. Costing is wrong by the same four percent, which makes the product look more profitable than it is — and because the error is proportional, it is largest on the highest-volume lines.
Setting the scrap factor to four percent fixes three things at once. Planning orders 1,040 sheets rather than 1,000, so the emergency purchases stop. Standard cost picks up the real consumption, so margin becomes accurate. And the stock adjustment that was absorbing the difference disappears, which means the next unexplained variance is genuinely a new problem rather than the same one again.
| Term | What it means |
|---|---|
| Parent / child | The assembly and the components that go into it. A multi-level BOM is simply a child that is itself a parent. |
| Phantom assembly | A sub-assembly that exists for structure but is never stocked — it is consumed straight into the parent. Useful for clarity, confusing if nobody documents why it is there. |
| Scrap factor | The expected loss per component, so planning orders more than the theoretical quantity. Getting this wrong is a quiet, permanent source of shortages. |
| Effectivity date | The date range over which a BOM version applies, which is how an engineering change takes effect without rewriting history. |
| Where-used | The reverse lookup: every parent a component appears in. The first thing you need when a supplier discontinues a part or a batch is recalled. |
The drawing is updated, the BOM is not. This is the single most common source of BOM drift.
A BOM with no scrap allowance under-issues components and overstates margin on every order.
Glue and packing tape in the BOM create transaction noise for no costing benefit. Consider overhead instead.
BOMs decay. A scheduled review against actual issues catches drift before stock valuation does.
| Metric | What it tells you | What distorts it |
|---|---|---|
| BOM accuracy | The share of BOMs matching what is actually consumed. Below 98% and material planning cannot be trusted. | Checking against drawings rather than against actual issues. |
| Component shortage incidents | How often production stops for a part that planning thought it had. | Buffer stock hiding the shortage rather than preventing it. |
| Average BOM depth | How many levels deep the structure runs. Deep structures are harder to cost and slower to plan. | Phantom levels inflating the count without adding real structure. |
| Engineering change lead time | Days from an approved change to it being live in production. | Measuring from release rather than from approval, which hides the queue. |
We build and support the systems this page describes. Start with a free working proof of concept — no cost, no commitment.