Cost per unit produced is the total cost of making one good unit, direct material, direct labor, and applied overhead (machine time, energy, tooling, and facility cost) divided by the count of sellable units. Scrap and equipment losses raise it by shrinking the denominator while the costs stay put.

Most plants carry a “standard” unit cost set once a year and quoted to customers. The number that runs the floor is different: it is the actual cost of the units that shipped this week, and it moves every shift with yield, downtime, and speed. This guide builds a true unit cost from the four buckets that feed it, shows how OEE losses inflate the figure invisibly, and works a hand-checkable example. If you want the loss side quantified first, the free OEE calculator gives you the denominator this whole calculation depends on.

What is cost per unit produced?

Cost per unit produced is total production cost for a period divided by the good units that came out of that period. The formula is simple; the honesty is in the two halves. On top you sum every cost consumed to run, material, labor, and the overhead that keeps the line moving. On the bottom you count only units you can sell. A part that gets scrapped still consumed its material and its share of labor and machine time, so its cost does not disappear; it reloads onto the good units that survived.

That is the trap in the metric. Two lines can buy identical material at the identical price and pay operators the identical wage, and still post unit costs that differ by 20%, entirely because one line converts more of what it starts into sellable product. Unit cost is a yield-and-effectiveness number wearing an accounting costume.

The four buckets that build a unit costWhat builds a cost per good unitDirect materialDirect laborMachine time +applied overheadScrap + rework(loaded onto survivors)÷GOOD units only= cost per unit producedtotal units are not the denominatorCosts sit above the line; only sellable units sit below it.
A true unit cost sums four cost buckets and divides by good units. Scrap does not lower cost, it moves the scrapped part’s cost onto the units that survived.

What goes into a true unit cost?

A true unit cost has four buckets, and the fourth is the one standard costing usually buries. Keep them separate so you can see which one is moving:

BucketWhat it coversMoves with
Direct materialRaw material, components, packaging that end up in the productPurchase price, usage, material yield
Direct laborOperator hours paid to run and change over the lineWage, crew size, run rate, downtime
Machine time & overheadEnergy, depreciation, tooling, maintenance, facility cost applied per run-hourUptime, speed, throughput
Scrap & reworkMaterial and time lost to defects, plus the labor to rework recoverable unitsFirst-pass quality, defect rate

The first bucket is mostly a purchasing story. The other three are effectiveness stories: they are fixed or semi-fixed per hour, so what determines their cost per unit is how many good units that hour produced. That is why unit cost and OEE are the same conversation from two directions, and why a plant chasing unit cost with only a purchasing lever is fighting with one hand.

How do OEE losses inflate cost per unit?

OEE losses inflate unit cost by shrinking the denominator while the fixed costs above the line barely move. A crew is paid the same for an eight-hour shift whether the line runs at 85% OEE or 55%. The building draws the same power for lighting and HVAC. Depreciation does not pause during a breakdown. So when downtime slow cycles, and scrap take a third of your output, roughly a third more fixed cost piles onto every unit that makes it out the door.

Work an example. Say fixed labor and overhead for a shift total \$4,000. At the line’s nameplate rate the shift would yield 10,000 good units, so fixed cost per unit is \$0.40. Run the same shift at 60% OEE and you get 6,000 good units, the same \$4,000 now spreads to about \$0.67 per unit. The material cost per unit did not change. The wage did not change. Unit cost rose 67% on the fixed portion, and nothing on a purchase order explains it. It hides inside the six big losses.

Fixed cost per unit rises as OEE fallsSame fixed cost, fewer units, unit cost climbs$0.4490% OEE$0.5080% OEE$0.5770% OEE$0.6760% OEE$0.8050% OEEIllustrative: $4,000 shift fixed cost, 10,000-unit nameplate output.
The fixed portion of unit cost is a mirror of OEE. Recover effectiveness and the same overhead spreads over more units, no price negotiation required.

How do you calculate cost per unit produced, step by step?

Pick a period short enough to act on, a shift or a run, not a quarter, and work these steps in order:

  1. Fix the period and pull the good-unit count. Count only first-pass good units. Reworked units cost more, not less, so do not quietly credit them as if they came out clean.
  2. Total direct material. Material issued to the run at standard or actual price, including the material embedded in scrapped parts. Do not net out scrap here; you want to see it.
  3. Total direct labor. Crewed hours × loaded wage for the period, including changeover and cleanup time, not just clean run time.
  4. Apply machine time and overhead. Use a per-run-hour rate that carries energy, depreciation, tooling, and maintenance. Multiply by the hours the line was staffed, because that is what you actually paid for.
  5. Add the scrap and rework load. Material and labor consumed by defects that produced nothing sellable, plus the extra labor to rework recoverable units.
  6. Divide and decompose. Sum the four buckets, divide by good units, then split the result into per-unit material, labor, overhead, and scrap so you know which bucket to attack. Recompute every period and watch the trend, not the single number.

What does a worked example look like?

These numbers are hypothetical chosen so you can check the arithmetic by hand. One shift on a packaging line:

InputValue
Good units produced (first pass)6,000
Direct material (incl. scrapped material)$3,000
Direct labor (crew × hours × wage)$1,600
Machine time + applied overhead$2,400
Scrap + rework load$500
Total cost$7,500

Cost per unit produced = \$7,500 ÷ 6,000 = \$1.25. Decomposed, that is \$0.50 material, about \$0.27 labor, \$0.40 overhead, and about \$0.08 scrap. Now the lesson: if this line had run at nameplate and produced 8,000 good units on the same staffed shift, labor and overhead (\$4,000 combined) would have spread over 8,000 units instead of 6,000. Fixed cost per unit falls from about \$0.67 to \$0.50, and total unit cost drops toward \$1.08, a 14% cut with no new supplier, no wage change, and no capital. The lever was effectiveness, tracked through throughput.

Why does standard cost hide the real number?

Standard cost is a budget: a per-unit figure set from expected material prices and an assumed run rate, used to quote jobs and value inventory. It is useful for planning and useless for improvement, because it bakes in an assumed OEE and then stops moving. When actual OEE drops below the assumption, the extra cost shows up later as an unfavorable variance on a monthly report, long after the shift where you could have done something about it.

The gap between standard and actual is the cost of poor effectiveness and it is real money. A plant that reports only standard cost is flying on a number that was true last January. Measuring actual cost per unit each shift, or better, computing it live from the same machine signals that feed OEE, turns unit cost from a backward-looking variance into a scoreboard the crew can move today. That is the same argument behind measuring at the source rather than reconstructing it from memory at month-end (see the platform).

Standard cost versus actual cost per unitThe gap standard cost hidesStandard (planned)$1.08Actual (this run)$1.25$0.17 gapThe effectiveness variance, visible now, or as a month-end surprise.
Standard cost assumes an OEE and stops moving. The gap to actual cost is the effectiveness variance, the same dollars, surfaced live instead of at month-end.

How do you drive cost per unit down?

Attack the bucket that is actually moving, in this order of usual payoff:

Track the components against a baseline the way a plant tracks its other manufacturing KPIs and pair the number with the loss detail behind it, the six big losses tell you which bucket is bleeding. See how one plant tied unit economics to floor effectiveness in the CLS case study.

Data & sources

The link between cost and effectiveness is not a Harmony idea; it is how national statistics agencies define productivity and cost.