Costing systems: what is hidden inside the unit cost your plant reports

What the cost roll assumes without saying so

Rolling standards up a bill combines a material price, a run rate, a setup allowance divided by an assumed batch size, and an overhead rate applied on some basis. Each is an assumption presented afterwards as a fact. The batch size assumption is the most consequential and least examined: a product quoted on a large batch and habitually made in small ones carries a setup cost the standard never captured. Documenting the assumptions alongside the standard, and revisiting them when the mix moves, is what separates a costing model from a number that gets defended because it is in the system.

Absorption changes behaviour before it reports anything

How overhead is applied is not a neutral accounting choice. Absorbing on labour time makes automated processes look cheap and manual ones expensive, which pushes work towards machines regardless of what the cash actually does. Absorbing on machine time does the reverse. A single plant-wide rate makes complex low-volume products look like simple high-volume ones. Departmental or cell-level rates take more effort and describe reality better. Whichever basis you use, somebody should be able to explain in plain language why a product carrying more overhead genuinely consumes more of it.

Variances are a diagnostic, not a scoreboard

Purchase price, material usage, labour rate and efficiency variances each point at a different mechanism, and together they explain why the period differed from the plan. Used that way they are the most useful output the system produces. Used as a performance measure for named managers, they become something to manage rather than something to read: purchasing buys in bulk to beat a price variance and inflates stock, production runs long batches to absorb overhead and blocks the schedule. The variance report survives; the information in it does not.

Actual costing needs discipline the floor may not have

Costing at actual removes the standards argument and replaces it with dependence on shop floor data. Every unreported scrap, mis-booked hour and undocumented substitution flows straight into the cost of a specific order. Where booking is sound this gives genuinely useful order-level truth. Where it is not, the plant gets volatile costs that nobody trusts and which vary more with reporting behaviour than with what happened. Assess booking quality honestly before choosing this route, because the accounting change does not create the discipline it depends on.

The decisions a distorted cost corrupts

Unit cost drives make-or-buy, product rationalisation, quoting and investment cases. When overhead is averaged across a mixed portfolio, low-volume complex products are subsidised by simple high-volume ones, so the business drops the profitable simple work as low margin and keeps the complex work that is losing money. Outsourcing decisions made on a standard including fixed overhead that will not disappear with the work are the other classic error. Before any of those decisions, ask which costs actually change if the answer changes, because that is a different figure from the one the system reports.

Frequently asked questions

Should we use standard or actual costing?

Standard costing suits repetitive production with stable processes, because the variance analysis it produces is informative and the transaction load is light. Actual costing suits project and job work where every order differs and an average would be meaningless. Many plants run standards for stock valuation and actual capture for specific orders where the answer matters commercially. The deciding factor is usually the reliability of shop floor booking rather than a preference in principle.

How often should standards be re-rolled?

Often enough that they still describe the process and rarely enough that variance analysis remains comparable across periods. Rolling constantly destroys the ability to see a trend, because the baseline keeps moving. A defined annual roll with controlled interim updates for material price movements that are large and permanent is a common compromise. What matters more than frequency is that a roll is triggered by process changes and mix shifts rather than only by the calendar.

Why does our costed margin differ from the financial result?

Because absorbed overhead is not actual overhead. If volumes ran below the level assumed in the rate, cost was under-absorbed and the difference appears in the accounts but not in any product margin. Add unreported scrap, stock adjustments, rework not booked to an order, and purchase price differences held centrally, and the gap becomes explicable. Reconciling the two figures once, line by line, is usually more instructive than any single costing refinement.