In this guide
Reconciling BOM to Shop-Floor Consumption
The bill of materials states what each unit of output should have consumed. The stores issue records state what was actually drawn from the floor. Multiply the output for a period by the standard consumption, compare it to the issues, and the difference is the question every component manufacturer has to be able to answer. Some of that difference is expected and allowed for: process loss, machining allowance, setting-up scrap and rejection at inspection are all normal, and a properly maintained BOM carries a standard allowance for them. What matters is whether the actual difference sits inside the allowance and stays there. A gap that widens without a change in product or process points somewhere specific: material issued and not returned, scrap generated and not recorded, output under-reported, or a standard that was set once and never revisited as the process changed. An auditor treats a persistent unexplained variance as a control finding, not a costing curiosity.
Reading the Bill of Materials
A bill of materials states the standard quantity of each input required per unit of output, and reading it correctly requires knowing what has already been built into that standard. The quantity is not the theoretical minimum. It generally includes an allowance for the material lost in the process itself: the swarf from machining, the offcuts from cutting, the setting-up pieces at the start of a run, and the proportion rejected at inspection. Where that allowance is included, comparing issues against a theoretical consumption will show a variance that is entirely expected. Where it is excluded, the comparison has to add it back. Knowing which convention the bill follows is the first thing to establish. BOM versions and which one applied is the second and it is the source of most spurious variances. Bills are revised when a design changes, a supplier changes, or a process is improved, and production during a period may have run under more than one version. A comparison performed against the current bill, over a period in which an earlier version was in force, measures the revision rather than the consumption, and the effect can easily exceed the leakage anybody is looking for.
Actual Issues From Stores
The other side of the comparison is what stores actually issued, and it is a record with characteristic weaknesses. Issue slips against production output is the basic test: material drawn for a job set against what that job produced. It works only where issues are booked against a specific work order, and where they are booked to a general cost centre the comparison cannot be made at the job level at all. Over-issue and unreturned balance is the largest distortion. Material drawn in a round quantity because that is how it is packed, with the remainder physically returned to the store but never credited in the system, shows as consumption that never happened. Every such issue inflates the variance permanently. Where returns are not documented at all, the whole comparison becomes uninterpretable. Issues to the wrong work order are the third weakness and they are self-cancelling in aggregate but not by job. Material booked to one order and consumed on another produces an overstated variance on the first and an understated one on the second, which is why apparent leakage on a single job frequently disappears when a group of related jobs is examined together.
Explaining the Variance
The gap between standard and actual is worked through in a fixed order, because each explanation has to be exhausted before the next becomes plausible. Normal process loss is the first and largest component: whatever the process inherently consumes, measured against the allowance the standard carries. Where the actual sits inside the allowance the variance is explained and no further work is needed. Rejection and rework consumption is the second layer. Units rejected at inspection consumed material that produced nothing, and units reworked consumed material twice, so both belong in the reconciliation and neither appears in a standard consumption computed from good output alone. Reconciling to output including rejects, rather than to saleable output, removes a large part of what looks like leakage. Where the residue points to leakage is the conclusion reached last rather than first. A gap surviving the process allowance, the rejection analysis and the returns adjustment, persisting across periods, and not explained by a bill revision, is material that left without a record, and it is investigated as a control matter rather than reported as a costing variance, because no adjustment to the standard will make it go away.
Evidence the Auditor Wants
Three records are needed to make the comparison meaningful, and a gap in any one leaves the variance unexplainable. Production output records come first, because standard consumption is computed from output and an output figure that is itself unreliable makes the whole calculation circular. Output has to be evidenced from the production records and tied to what was actually transferred to finished goods, not taken from a plan. Return-to-stores documentation is second and is the record most often missing. Material drawn in excess of what a job needed and physically returned to the store, but never credited back in the system, shows up as consumption that never happened and inflates every variance computed afterwards. Where returns are undocumented the variance cannot be interpreted at all. Third is scrap generated against scrap expected, since scrap is the legitimate destination for the difference between issue and consumption. Recorded scrap materially below the process norm points to material leaving without record, and recorded scrap materially above it points to good material being disposed of, and both are established from the same comparison.
Running the Reconciliation Yourself
Select part numbers by value and by volume rather than attempting the whole bill of materials, because the exercise is diagnostic and a complete reconciliation costs far more than it reveals. High-value materials show what the money is doing; high-volume materials show whether the process discipline holds, since a small percentage variance on a part issued in tens of thousands is a large absolute quantity. Between them, a few dozen part numbers will tell you most of what a full reconciliation would. Period selection and cut-off matter more than people expect. Choose a period long enough to average out normal batch variation, usually a quarter rather than a month, and make sure the output figure and the issue figure cover exactly the same window. A mismatched cut-off produces a variance that is entirely artificial, and it is the most common reason a first attempt at this exercise produces alarming results. Independent reconciliation is needed where the variance persists without explanation, where the standards themselves are suspected of being stale, or where the finding will go to a lender, and stock audit for auto components covers exactly this ground.
