Costing a Finished Good: From BOM to a Price That Holds
Manufacturers price from a bill of materials and lose money on the difference between the recipe and the run. Where the real unit cost comes from, why yield belongs in the price rather than in the variance, and the review rhythm that keeps a price list honest while input costs move.
A small manufacturer prices a product by adding up the bill of materials, adding something for labour, and applying a margin. It is a reasonable method and it produces a number that is consistently too low — usually by more than the margin itself, which is why manufacturers can be busy, growing and unprofitable at the same time.
The gap is not in the arithmetic. It is that the bill of materials describes a perfect run, and no run is perfect.
Four things the BOM does not tell you
| Missing from the recipe | What it does to unit cost |
|---|---|
| Yield loss | You buy more input per unit of output than the recipe states — sometimes materially more |
| Landed cost on imported inputs | Freight, duty, clearing and handling are real and are routinely omitted |
| Rework and rejects | A unit produced twice consumed inputs twice and sells once |
| Idle and setup time | Labour paid during changeovers is a genuine cost of producing a mix |
Yield is the largest of the four in most operations and the most misunderstood. It is not waste in the sense of carelessness — it is trimming, evaporation, spillage, off-cuts, the tail of a batch, and the material that stays in the machine. It happens on every run, it is predictable, and it belongs in the standard cost rather than being discovered as a variance.
Yield loss is not carelessness. It is trim, evaporation, spillage and the material that stays in the machine — predictable, unavoidable, and missing from every price built straight off a bill of materials.
Cost the run, then price the unit
The correction is to derive unit cost from actual production runs rather than from the recipe, and to update it as inputs change. The arithmetic is simple and the discipline of doing it regularly is what separates manufacturers who hold their margin from those who discover its absence annually.
One product, recipe cost against run cost
Illustrative, in KES. The manufacturer believes they are earning 35% and is losing 28 shillings a unit — and because volume is growing, the loss grows with success. Every line added after the BOM is real, ordinary and individually modest; together they are 50% of the recipe cost.
Standard cost with a periodic reality check
Costing every unit from actual consumption is impractical for a small manufacturer. The workable approach is a standard cost per product, reviewed on a rhythm, with a reality check against actual runs.
-
Set a standard from three or four real runs
Not from the recipe and not from one exceptional run. Three runs establish the yield and rework you actually get rather than the ones you intend.
-
Update input costs at receiving
Landed, not invoiced. The lag between a cost rise and a price response is where margin quietly disappears — see landed costs in Kenya.
-
Re-derive the standard quarterly, or whenever an input moves sharply
A standard cost that is a year old is a historical document. In an import-dependent operation, a currency move alone can justify a re-derivation.
-
Compare standard against actual monthly
A persistent gap means the standard is wrong or something changed on the line. Both are worth knowing; neither is visible from a monthly production total.
Contribution, not full cost, for pricing decisions
For deciding whether to accept an order or add a product, direct cost is the number that matters — materials, yield, rework and direct labour. Allocating factory overheads across products produces a figure that feels rigorous and misleads on marginal decisions, because the overhead is there whether or not you take the order.
Price lists move slower than costs
The second structural problem is timing. Input costs change continuously; price lists change when somebody decides to change them, which in most small manufacturers is once or twice a year and always later than it should be.
The practical defence is a scheduled review rather than a reaction: a fixed slot each quarter where standard costs are re-derived and prices are examined together, with the margin exception list — items now below cost or at thin margin — in front of you. Batching the decision produces better outcomes than reacting to individual supplier letters, and it gives the task an owner.
One market-specific note: where a large share of your inputs is imported, the shilling is an input cost like any other. A manufacturer whose prices have not moved through a significant currency shift has taken the whole of it out of margin, usually without deciding to.
What we do and do not do
What AWRA OpsHub does today
- Bills of materials defining the inputs to a product.
- Landed cost components added to input costs, so the material figure reflects what it cost to get here.
- Stock issues and receipts against production, so actual consumption per run is recorded.
- Item cost history, which is what makes a quarterly re-derivation possible.
- Margin signals — items below cost, at thin margin, or with no price — surfaced as an exception list.
- Job costing, so a production run can carry its materials, time and expenses.
What it does not do
- No standard costing engine. There is no standard-versus-actual variance calculation; the comparison described here is an analysis you run.
- No yield or scrap factor on a BOM. Yield is applied by you when deriving cost — the bill of materials states inputs, not expected loss.
- No overhead absorption. Factory overheads are not allocated to products by any rule.
- Not an MES. No machine integration, no production scheduling, no capacity planning, no shop-floor data capture.
The absence of a yield factor on the BOM is the one that matters most for pricing, so build it into how you derive the standard rather than expecting the recipe to carry it. The yield discipline itself is covered in raw materials and yields.
Our take
Derive your standard cost from three real runs rather than from the recipe, add landed cost and yield explicitly, and re-derive quarterly with the margin exception list in front of you. Price on direct contribution for marginal decisions. A manufacturer growing on a product priced from the BOM is scaling a loss, and volume makes it worse rather than better.
See landed costs and production costing
Bills of materials, landed cost on inputs, consumption recorded per run, item cost history and margin exception signals.
Explore costingFrequently asked questions
Why is our BOM cost lower than what production actually consumes?
Because a bill of materials describes a perfect run. Yield loss — trim, evaporation, spillage, the tail of the batch, material left in the machine — is real on every run and is not in the recipe, and neither is rework. Add landed costs on imported inputs and setup labour and the gap can reach half the recipe cost. None of it is carelessness; it is simply missing from the document you priced from.
Does the system apply a yield factor to a BOM?
No — the bill of materials states inputs, not expected loss, and there is no scrap or yield factor field. You apply yield when deriving your standard cost, which is why deriving it from three or four actual runs rather than from the recipe matters. Actual consumption per run is recorded, so the data to derive it is there; the calculation is yours.
Should we allocate factory overheads to products?
Not for pricing and marginal decisions. Direct cost — materials, yield, rework, direct labour — is what tells you whether an order or a product is worth taking, because the overhead exists whether or not you take it. Full absorption costing has its place in financial reporting and consistently misleads on the operational decisions manufacturers actually face. Note that nothing here allocates overheads automatically in any case.
How often should we re-derive standard costs?
Quarterly as a rhythm, and immediately whenever an input moves sharply — a currency shift alone can justify it in an import-dependent operation. A standard cost more than a year old is a historical document, and pricing from it means you are quoting last year's economics. Pair the re-derivation with a scheduled pricing review so the two decisions happen together rather than months apart.
What is the fastest way to find products that are underpriced?
The margin exception list — items now selling below cost, at unusually thin margin, or with no price set. It is a fifteen-minute read and it surfaces exactly the lines where an input cost moved and the price did not follow. It will also flag deliberate loss-leaders, which is a small annoyance worth tolerating: reviewing the list monthly is how the accidental ones get found while they are still small.
