True Cost Per Batch in Manufacturing
Search for batch costing and you get the same formula several times over. Materials plus labour plus overhead, divided by quantity produced. It is correct, and it is not the hard part.
The formula has two inputs that look obvious and are not: which material price, and which quantity. Getting either wrong produces a confident number that is quietly too low, and too low is the dangerous direction, because you price off it.
Which price: the average or the lot
Most systems carry one cost per product, updated periodically. It is an average of what you have been paying, and it is smooth.
Smooth is the problem. If your supplier raised the price of a key ingredient eight weeks ago, an average tells you your cost drifted up a little across everything. Costing against the specific lot the batch consumed tells you that this batch, the one you are pricing, cost what it cost.
Two runs of the same product a month apart can then come out at genuinely different costs, which looks like noise and is not. It is the supplier price change arriving at the moment it actually arrived. That is also why which lot gets picked has a cost consequence and not only a freshness one.
Which quantity: the plan or the pallet
The denominator does more damage than the numerator.
You planned two hundred jars. You filled a hundred and eighty, because the first ten minutes of the run were off spec and one carton went over on the floor. The material for two hundred is gone either way. Divide by two hundred and every jar looks eleven percent cheaper than it was.
This is not an accounting subtlety. It is the difference between a product you think earns thirty percent and one that earns nineteen, and it compounds because the same optimistic number goes into the next quote.
The costs that get left out
Scrap and wastage. Material lost in production was bought. A ten percent scrap rate means paying for a hundred and ten units of material to ship a hundred. If wastage is not recorded against the batch it silently becomes a shrinkage number somewhere else, disconnected from the product that caused it.
Rework. A batch that failed inspection and was reworked consumed labour twice and sometimes material twice. Recording only the successful pass makes the failure free, which is exactly backwards from what you want the number to tell you.
Intermediates. When a batch is consumed into another batch, its cost has to travel with it. If the second run starts its costing from scratch, the finished good carries the price of the intermediate as though it appeared from nowhere.
What this looks like when the system holds it
In IEMSuite the material side of a batch cost comes from the specific lots consumed rather than a stale average, so it reflects what you really paid, and recorded wastage is included. Labour attributed to the batch and overhead allocated to the run roll up into a total, and the total divided by the batch output gives the per unit figure. How production cost is calculated covers where each part comes from, and recording wastage is what keeps the scrap attached to the batch that produced it.
What no system decides for you is how to allocate overhead, or what counts as labour on a run where three people drifted in and out. Those are your conventions. The useful thing software does is stop the two inputs above from being guesses.
A test you can run this week
Take your last completed batch. Write down the cost per unit your current method gives. Then recompute it with the actual price of the lots that run consumed, and divide by what came off the line rather than what the recipe said.
If the two numbers agree, your costing is doing its job. If the second is meaningfully higher, that gap has been in every quote you have sent.
Questions people ask
How do you calculate cost per batch?
Add the material consumed, the labour attributed to the run and the overhead allocated to it, then divide by what the batch actually produced. The formula is not where the number goes wrong. Which material price you used, and whether you divided by theoretical or actual output, is.
Should I divide by theoretical or actual yield?
Actual. If you planned two hundred units and got a hundred and eighty, the material for all two hundred was still consumed, so dividing by the plan understates unit cost by more than ten percent. The gap between planned and actual output is a cost, and hiding it in the denominator is the most common way a batch looks cheaper than it was.
Which price should I use for a raw material?
The price you paid for the specific lot the batch consumed. An average across everything on hand smooths out exactly the movement you need to see, which is what happens to margin when a supplier raises a price. If two batches of the same product cost different amounts because they drew from different lots, that difference is real.
Does scrap belong in the batch cost?
Yes, and it usually is not counted. Material lost in production was bought and paid for. A ten percent scrap rate means buying a hundred and ten units worth of material to end up with a hundred good ones, and if that is not in the number then the number is describing a batch that did not happen.
See it in a real system
IEMSuite is lot-tracked MRP for small manufacturers. Free during Early Access.
Start free