Warehouse KPI formulas you can check.

Start with a shared definition.

A warehouse KPI only helps when the team agrees what the number includes. Before comparing this week with last week, write down the building, date window, unit of work, labor scope and exclusions. Keep the definitions beside the report so a colleague can reproduce it.

The four formulas below are the working definitions used for DC Health Check’s scoped reporting. They are not universal industry standards or target benchmarks. Adapt them to the records you actually have, then keep the chosen definitions consistent.

1. Labor cost per completed shipped order

Choose the cost scope first. Document whether wages, agency fees, employer costs and overtime premiums are included. Match the labor to the building and reporting period used for shipments. A warehouse-wide payroll total cannot automatically describe only the packing process.

Data to collect: completed shipment date, a stable order identifier, shipment status, and daily labor costs with the agreed scope. If an order appears on several shipment or item rows, reconcile those rows to the agreed definition of a completed order before counting.

Question to investigate: did the cost change because paid hours rose, the cost per hour changed, shipped volume fell, or the mix of work changed? The ratio alone cannot establish a cause. A week spent building inventory or clearing difficult backlog can change the comparison.

2. Units per direct paid hour

Use units in the numerator throughout the report. Orders, order lines, units and cartons describe different amounts of work. The example’s 15 units per hour cannot be compared directly with a report showing orders per hour.

Paid hours and productive hours are also different denominators. A paid-hour measure can include paid breaks and other paid time within the agreed labor scope. A productive-hour rate used in a staffing model excludes the time you have defined as nonproductive. Do not move between them without reconciling the hours.

Data to collect: completed shipped quantity, units of measure, daily direct paid hours and the activities included. Compare similar order mix and process scope. If the mix changed, label that as a limit on the comparison.

Using the rate for peak planning? The free warehouse capacity calculator needs orders per productive labor hour for one process. It should not be populated with the units-per-paid-hour example above.

3. On-time shipment rate among completed orders

Define “shipped” using an available event such as the agreed completed-shipment timestamp. Compare it with the documented promised ship time for that same order. A ship promise is different from a delivery promise; warehouse shipment records do not necessarily establish delivery performance.

Use the same time zone and business rules on both timestamps. Record whether equality at the promised time counts as on time. The example treats a shipment at or before its promise as on time.

Make the coverage visible: if 200 of 4,000 completed orders lack a valid promise, report 3,800 eligible orders and 200 excluded records. Do not silently treat the missing promises as passes or failures. The worked example above has no such exclusions.

4. Recorded overtime share

The overtime hours must be a subset of the paid hours in the denominator. Use the overtime hours explicitly recorded in your accepted labor export. Do not estimate them from a long day or add them to paid hours a second time.

This measure describes the share of hours recorded as overtime. It is not the share of labor cost spent on overtime, and it does not determine payroll or legal treatment. In the example, 48 of 640 paid hours are recorded overtime, so the share is 7.5%.

Five checks before sharing the report

• Use matching periods. Compare complete windows with the same weekday mix. Label holidays, shutdowns and material workload changes.
• Reconcile duplicate records. Decide what one row represents before counting orders or summing hours. Keep excluded records traceable.
• Calculate from totals. For a period ratio, divide the summed numerator by the summed denominator. A simple average of daily percentages can give a different answer when daily volume varies.
• Show unavailable values honestly. A zero or missing denominator produces an unavailable KPI, not zero performance. Missing hours, costs or promises need a visible exception.
• Separate evidence from the next check. “Cost per order rose” is an observation. “Check the shift’s paid-hour mix” is a proposed investigation, not a proven cause.

What to include in shipment and labor exports

For a first discussion, start with field names and a question. A useful shipment export can include order ID, completed-shipment date/time, shipped units, status and promised-ship date/time where available. A labor export can include date, agreed direct paid hours, recorded overtime hours and agreed labor cost.

Document the building, time zone, units and row structure. Aggregate labor by day where that fits the question. Remove employee names, customer details, payment information and credentials before sharing any sample, and agree the transfer method first.

You do not need every field to ask a useful question. The available fields determine which measures can be supported. Use the duplicate-order checklist if your export repeats order IDs and you are unsure what to count.