Online Tool Store
· 4 min read
How to Allocate Overhead Across Products
Heshan Fernando
Co-founder & COO
Product A looks profitable and product C looks marginal. Someone suggests discontinuing C. Then somebody re-runs the numbers allocating overhead on machine hours instead of labour hours, and C is the better product.
Nothing about the business changed. Only the divisor did — and that’s worth understanding before anyone discontinues anything.
What allocation is doing
Overhead is cost that isn’t directly traceable to a unit: rent, supervision, machine depreciation, quality control. To report a cost per product, you have to spread it somehow.
An absorption rate does that in proportion to a chosen driver.
rate = total overhead ÷ total driver units
allocated = rate × that product's driver units
With 120,000 of overhead and 8,000 labour hours, the rate is 15.00 per hour. A product consuming 4,000 hours absorbs 60,000.
The assumption buried in that is causation: that the driver is what causes the cost. Sometimes it is — machine-related overhead genuinely follows machine hours. Often it isn’t. Setup costs, handling and quality inspection follow the number of batches, not the number of units, which means volume-based allocation systematically under-charges small complex batches and over-charges long simple runs.
That’s the classic cross-subsidy in absorption costing, and it’s the argument for activity-based costing — allocating each pool of overhead on the activity that actually drives it.
Why people get stuck here
- One driver for everything. A single plant-wide rate applied to overhead with several different causes.
- The driver picked by convenience. Labour hours are measured already, so labour hours get used, whether or not they cause the cost.
- Allocation mistaken for reality. The reported per-unit cost feels like a fact, when part of it is a choice.
- Decisions from allocated costs. Discontinuing a product removes its revenue and its direct costs — but the overhead mostly stays and gets reallocated to what’s left.
What good allocation looks like
Multiple pools where causes differ
Split overhead into pools with distinct drivers: machine costs on machine hours, setup costs on batch counts, handling on movements. Two or three pools capture most of the benefit.
Sensitivity tested
Run the allocation with a second driver. If product ranking is stable, the choice doesn’t matter much. If it flips, the driver is doing the analytical work, and that needs saying out loud.
Kept separate from decisions
Allocation is for reporting and pricing context. For a keep-or-drop decision, look at contribution — revenue minus the costs that would actually disappear.
| Overhead Type | Sensible Driver | Poor Driver |
|---|---|---|
| Machine depreciation, power | Machine hours | Units produced |
| Setup, changeover | Number of batches | Direct labour hours |
| Materials handling | Number of movements | Material cost |
Common mistakes to avoid
- Using one plant-wide rate for overhead with genuinely different causes.
- Treating allocated cost as avoidable when deciding whether to drop a product.
- Allocating on direct labour in a facility that’s mostly automated — a hangover from when labour was the dominant cost.
- Changing driver mid-year, making period comparisons meaningless.
- Presenting a per-unit cost without noting which driver produced it.
How to do it with Overhead Allocation Calculator
The Overhead Allocation Calculator computes the rate and the allocation, and makes it easy to try a second driver.
- Enter total overhead for the period and the driver you’re allocating on.
- Enter each product’s share of that driver.
- Read the absorption rate and each product’s allocated cost.
- Run it again with a different driver and compare — that comparison is the real output.
Other costing and pricing tools are in the tools directory.
Frequently asked questions
Which driver should I choose?
The one that causes the cost. Machine-heavy overhead follows machine hours; setup and handling follow batch counts rather than volume. If one driver can’t sensibly explain all your overhead, split it into pools.
Why does my low-volume product look profitable?
Volume-based allocation under-charges small, complex batches because they consume setup and handling out of proportion to their unit count. That cross-subsidy is the standard weakness of a single-driver absorption rate.
Does allocation change total profit?
No. It only moves cost between products. It changes which products look profitable, which is exactly why the driver choice deserves scrutiny.
Final thought
Run two drivers before anyone makes a decision. The gap between the results tells you how much of your product profitability is a business fact and how much is an accounting choice.