Supplier Relationship Management, Performance Measurement, and Continuous Improvement
This domain tests governance: how you segment suppliers, measure them without being gamed, and drive improvement that sticks. Expect questions that contrast a well-designed scorecard with a vanity dashboard, and that ask which improvement tool fits which problem.
Segment the relationship before you manage it
Relationship intensity should follow Kraljic position: transactional/arm's-length for non-critical, competitive for leverage, secured and developed for bottleneck, and partnered for strategic. Strategic relationships need an executive sponsor, a joint business plan, defined governance cadence (operational weekly, tactical monthly, executive quarterly), shared risk and gain-sharing terms. Supplier development is the right answer when switching cost is high and the supplier's capability gap is closable; switching is right when the gap is structural.
Measure what changes behaviour
- Build a metric hierarchy: business outcomes (perfect order, total cost), process metrics (cycle time, plan attainment), supplier metrics (PPM, OTIF, responsiveness, cost performance).
- SCOR groups performance into reliability, responsiveness, agility, cost and asset management - useful because it stops a scorecard from measuring only what is easy.
- Keep the scorecard small and weighted. Fewer than ten metrics, targets agreed with the supplier, and a consequence attached: award, volume shift, corrective action, or exit.
- Metrics get gamed. If you reward OTIF alone, suppliers ship early and incomplete; pair service metrics with cost and quality. This is Goodhart's law in a supplier review.
Continuous improvement
- PDCA is the governing loop; DMAIC structures a six-sigma problem; 8D or a SCAR structures supplier corrective action with containment, root cause and verification of effectiveness.
- Lean toolbox: seven wastes, value stream mapping, 5S, kanban/JIT, poka-yoke, setup reduction. Root cause via 5 Whys and fishbone - a corrective action without verified root cause recurs.
- Benchmark internally and externally, and re-audit after the fix; an improvement claim that is not verified is a claim, not a result.
Worked example
Scorecard weights: quality 40%, delivery 30%, cost 20%, responsiveness 10%. Scores: quality 3 (500 PPM), delivery 2 (92% OTIF), cost 4 (1% below target), responsiveness 4. Weighted = 1.2 + 0.6 + 0.8 + 0.4 = 3.0. That places the supplier in a conditional band: trigger an 8D for delivery, require a 90-day improvement plan with weekly OTIF reporting, and hold new volume until two consecutive months exceed 97%. Note that averaging hides the failure - a single metric at 2 on a critical dimension deserves its own escalation gate.
Traps that cost marks
- Scorecards with fifteen metrics and no consequence attached.
- Measuring supplier price performance while ignoring total cost and quality cost.
- Accepting a corrective action that fixes the symptom without root-cause verification.
- Applying lean tools to a problem that is actually a specification or design issue.
- Treating all suppliers to the same governance cadence regardless of segment.