Warehouse Reporting
Part of Fulfilment performance reporting
Reporting fulfilment errors per completed order
Define completed orders and verified defects, then report affected orders and error incidents with clear denominators.
Report the share of completed orders with a verified fulfilment error. Separately report the number of error incidents per completed order. One order can contain several errors. Define the completion event, in-scope defects and reporting period before comparing teams or weeks.
Fix the denominator
A completed-order cohort can include customer orders whose final required fulfilment part was physically handed over during the period. If the business uses a storefront’s fulfilled status instead, label the measure accordingly. Do not treat a system status as proof of physical handover unless it records that event.
Count each eligible customer order once. State how cancellations, collection orders and orders with mixed delivery methods are treated. Keep a split order open until its final required part reaches the chosen completion event. A replacement parcel for a correction should not quietly create another customer order in the denominator.
Steps to Accurately Report Fulfilment Errors
- Define completion event (e.g., physical handover)Yes
- Exclude cancellations and collection orders from denominatorYes
- Keep split orders open until final part is deliveredYes
- Do not count replacement parcels as new ordersYes
Verify and classify defects
Use categories that point to a correction: wrong item or variant, wrong quantity, missing required item, incorrect destination label, parcel linked to the wrong order, or a verified packing omission. Record whether the problem was found during picking, packing, dispatch review or after customer contact.
A problem caught and corrected before completion belongs in an internal rework measure. State explicitly if the published completed-order rate includes such cases.
A customer complaint is a lead, not automatic proof of a warehouse error. Check the order line, work and parcel records before verifying a defect. Keep cases under investigation visible but separate from the verified numerator.
For each verified case, retain the order ID, type, affected line or parcel, discovery and completion dates, evidence reference and correction. Two defects on one order mean two incidents and one affected order.
Calculate both measures
Affected-order rate = completed orders with at least one verified in-scope error ÷ eligible completed orders.
Incidents per 100 completed orders = verified in-scope incidents ÷ eligible completed orders × 100.
For a hypothetical 200-order cohort with five affected orders and six incidents, the affected-order rate is 2.5% and the incident frequency is three per 100 completed orders. These are arithmetic examples, not results from an operation.
Show the counts beside both rates. Separate defects caught before customer handover from those discovered later. More warehouse-detected cases can mean better detection, worse work or both; inspect the customer-reaching subset before drawing a quality conclusion.
Allow for later reports
Recent completion cohorts may still receive customer reports. Use a fixed observation window or mark newer cohorts provisional until they have had comparable time for defects to surface. When a late case is verified, update its original completion cohort under the stated rule and retain the earlier reported version.
Break results down by defect type or order pattern only where there are enough cases to inspect. Trace repeated items, locations or shifts back to source records before assigning a cause. Then change the relevant check or instruction and assess later cohorts with the same completion and discovery rules.
