Batch Yield and Unit Cost: Divide by Good Output

Batch cost per good unit uses saleable output as its denominator. Dividing by units started understates the cost when some units fail inspection or are lost during production. The yield assumption connects materials and labor consumed to the quantity the business can actually sell. It should come from measured production results or an explicitly labeled planning scenario, not an invented industry norm.

Worked calculation

Consider a batch that starts 1,000 units and incurs $4,800 of materials, $1,500 of direct labor, and $700 of other included production costs. Total batch cost is $7,000. If 900 units pass inspection, yield is 90% of the units started and cost per good unit is $7,000 divided by 900, or $7.7778. Dividing by the 1,000 starts would incorrectly present a $7 cost for each saleable unit.

Suppose 50 failed units can be reworked for an additional $200 and 40 of them then pass. Good output increases to 940 units and total batch cost rises to $7,200, giving approximately $7.6596 per good unit. The average unit cost falls, but the rework decision should also be evaluated incrementally. The extra 40 good units cost $200, or $5 each, before any further selling or fulfillment costs.

Decision checks and limits

Do not infer that every yield improvement reduces total cash spending immediately. A better pass rate can produce more output from the same paid labor shift and material batch. Its benefit may be additional units available for sale rather than a lower payroll payment. If demand is limited, the extra units may remain in inventory. Likewise, a process change that raises yield but adds substantial inspection labor needs a complete comparison of costs and feasible sales.

Keep first-pass yield distinct from final yield after rework. Track units started, initially accepted, sent to rework, recovered, and finally rejected so the totals reconcile without counting recovered units twice. Assign any scrap proceeds or supplier recovery consistently and show it separately. If a batch yields zero good units, cost per good unit is undefined; the batch still has a real total loss. Report that loss and the zero output instead of displaying a zero unit cost.

Apply the calculation

1. Reconcile physical output

   Use counts that distinguish initial failures from final rejects and identify the units recovered through rework.

2. Separate average and incremental cost

   Compare the cost of additional recovered units with the contribution they can produce, rather than relying only on the new batch average.

3. Retain failed-batch visibility

   Show total cost even when usable output is zero, and investigate the process before applying an average from successful batches.

Method references: OpenStax: Merchandising, manufacturing, and service costs. Accessed October 22, 2026. All dollar figures are hypothetical worked examples.