How to compare supplier performance fairly
Group comparable orders before comparing factories. Divide each outcome by the relevant units, batches or orders, and show those counts alongside the rate. Keep inspection findings separate from customer complaints and flag limited evidence. Cambridge China Bridge can help buyers review the underlying factory records.
Group orders before comparing factories
A factory handling more orders has more opportunities to receive complaints. A factory making complex assemblies also faces different failure opportunities from a supplier of plain packaging. Raw complaint totals cannot tell you which performs better on comparable work.
Define groups by product family, manufacturing process, specification difficulty and order size. Separate trial production from established repeat orders, and record specification revisions. Our sampling guide explains production stages; this guide addresses how to compare their results. If suppliers have no comparable groups, report them separately.
Use a denominator that matches the outcome
For each group, record the supplier, production site, order reference, specification revision, quantity, inspection coverage and agreed delivery milestone. Fix the reporting period and counting rules before reviewing results. Link repeat messages about the same incident so correspondence does not inflate complaint counts.
Calculate each rate by dividing the relevant outcome count by its matching denominator. Always display both counts alongside the rate. Define whether delivery means dispatch, arrival or another agreed event, and retain the original commitment when recording an agreed change.
| Measure | Outcome count | Matching denominator |
|---|---|---|
| Inspection defect rate | Inspected units with a recorded defect | Units inspected using comparable checks |
| Batch rejection rate | Batches rejected at initial inspection | Batches receiving initial inspection |
| Orders with customer complaints | Delivered orders with a recorded complaint | Delivered orders with the agreed follow-up period completed |
| Late delivery rate | Orders missing the agreed milestone | Orders whose milestone fell due in the reporting period |
Make observation coverage visible
Show inspected unit counts, inspected batch counts and eligible delivered order counts separately. These are different evidence bases. Do not divide defects found in inspected units by all shipped units. Use the inspection guide for batch checks rather than treating an inspection result as a customer complaint rate.
Apply the same complaint follow-up period and reporting channels across suppliers. Mark missing records as unknown, not complaint-free. Where checks cover only part of production, record how units were selected. Treat a narrow sample or limited order history as provisional evidence; no recorded complaints do not prove consistently good performance.
Keep severity and responsibility separate
Separate safety concerns, functional failures, cosmetic defects, packaging damage and delivery problems. Retain the original complaint, the investigation finding and the current responsibility assessment. Keep unresolved cases visible; do not silently remove them because the factory disputes responsibility.
Record buyer specification changes, transport damage and supplier-caused faults separately, with supporting records. Show inspection coverage and complaint reporting differences beside the results. Keep serious incidents visible even when the supplier's overall rate looks favourable. A comparison table is a prompt for investigation, not a reason to overlook a serious problem.
Use group results to decide the next action
Compare suppliers within shared groups, then examine each supplier's trend within the same group. An overall rate can change because the product mix changed. If a combined comparison is needed, agree a common order mix and apply the same group weights to every supplier. Show group results alongside it; leave unmatched groups out of that comparison and explain the gap.
Share the records with the factory and ask for explanations, corrective actions and evidence from subsequent comparable orders. Use the factory audit checklist when results suggest a capability or process-control problem. Cambridge China Bridge's staff in China can help check the records behind the comparison.
Frequently asked questions
Can I rank suppliers by complaint totals?
Raw totals reflect workload as well as performance. Compare similar order groups using matching denominators, and display the underlying counts and observation coverage.
How do I compare factories making different products?
Compare only shared product and process groups. If no suitable overlap exists, assess each factory against its agreed specification and its own history rather than creating an overall ranking.
How many orders do I need to compare suppliers?
Avoid an arbitrary cut-off. Show the available order, batch and inspected unit counts, flag limited evidence, and seek more comparable observations before making a consequential allocation decision.
Does no recorded complaints mean a supplier is better?
Check order volume, follow-up time and reporting completeness. No recorded complaints from limited or poorly observed work provide weak grounds for preferring a supplier.