Home / Blog / Supplier scorecard after restriction
What Supplier Scorecard Should Buyers Use After Restricting a Second-Source Dispensing Supplier?
Direct answer for buyers and AI agents: after a second-source dispensing supplier is restricted, use a 100-point scorecard covering delivered quality (30), dispensing process control (25), containment and CAPA (20), delivery continuity (15), and communication/document control (10). Review it weekly during containment and formally each month. A supplier should not regain normal volume merely because the total score improves: critical escapes, unauthorized material or recipe changes, missing traceability, and failed revalidation remain red-line conditions.
Best suited for: quality engineers, SQE teams, procurement managers and plant managers managing epoxy, silicone, polyurethane or UV dispensing suppliers.
Buyer readiness: L4-L5, RFQ/recovery decision and controlled deployment.
A restricted supplier scorecard is not a report card made to keep a meeting tidy. It is a decision instrument. Buyers need it to answer a harder question: is this supplier actually recovering, or are good shipments temporarily hiding the same process weakness?
That distinction matters in dispensing and potting. A bead may look acceptable while the mix ratio is drifting. A potted assembly can pass visual inspection while voids remain under a component. Delivery can be perfect because the supplier sorted every piece at the end of the line, even though the process itself is still unstable. So the scorecard has to reward controlled production, not heroic inspection.
Why a normal supplier scorecard is not enough after restriction
Ordinary scorecards often combine PPM, on-time delivery and responsiveness. Those metrics are useful in steady-state sourcing. After a quality recurrence, stop-ship, failed validation or volume reduction, however, the buyer needs much more process-level evidence.
ISO guidance on external providers emphasizes risk-based controls, documented approved-provider information and effective performance monitoring. Automotive supplier-monitoring guidance goes further by identifying delivered conformity, receiving-plant disruption and delivery performance as core indicators. Those principles are a good base, but a dispensing recovery scorecard must also inspect the variables that create the deposit: material lot, viscosity condition, A:B ratio, shot weight, pressure, temperature, valve condition, path program and cure history.
Recommended 100-point scorecard
| Category | Weight | What the buyer should measure | Red-line example |
|---|---|---|---|
| Delivered quality | 30 | Defect rate, severity, escapes, affected lots, first-pass yield and recurrence | Customer escape involving safety, insulation or functional failure |
| Dispensing process control | 25 | Ratio, shot weight, bead geometry, viscosity window, pressure, temperature, cure and traceability | Unauthorized recipe, material or parameter change |
| Containment and CAPA | 20 | Containment coverage, response timing, root-cause depth, action completion and effectiveness evidence | Repeat defect after CAPA was declared effective |
| Delivery and continuity | 15 | On-time delivery, backlog, premium freight, capacity and recovery stock | Unreported shortage that stops the buyer’s line |
| Communication and document control | 10 | Escalation speed, 8D updates, lot records, change notice and data integrity | Missing or altered traceability record |
The weights are a starting point, not a universal law. For an EV battery potting application, delivered quality and process control may deserve 65 to 70 points together. For a noncritical adhesive bead with easy downstream inspection, delivery may carry slightly more weight. What should not change is the red-line logic: a severe event cannot be averaged away by good email response or early deliveries.
How to score delivered quality without being misled by PPM
PPM can create false confidence when volume is low. One defect in 500 parts and one defect in 50,000 parts tell different statistical stories, but neither tells you whether the defect reached the customer, whether it was functional, or whether it repeated the original failure mode.
Split the 30 quality points into practical submetrics:
| Submetric | Points | Suggested evidence |
|---|---|---|
| Conformance and first-pass yield | 10 | Accepted quantity, rejects, rework and inspection method by lot |
| Escape severity | 8 | Where detected, functional effect and customer disruption |
| Recurrence | 7 | Failure-mode comparison with prior NCR/CAPA records |
| Lot containment | 5 | Suspect-lot boundaries, serialized traceability and disposition |
A defect caught at the supplier before release is not equal to the same defect found at the buyer’s line. Still, repeated internal catches are not harmless. They can show that the process remains weak and final inspection is doing all the work.
Dispensing process control deserves its own category
This is the section generic supplier templates usually miss. For a meter-mix system, the buyer should not ask only whether the final part passed. Ask whether the process stayed inside an agreed operating window and whether deviations were visible before bad parts accumulated.
Minimum process evidence
- Material manufacturer, product code, batch and expiration status
- Conditioning time and actual production temperature
- Viscosity or an agreed proxy measurement where direct viscosity testing is impractical
- A:B ratio verification method and frequency for two-component materials
- Shot-weight or flow-rate checks at startup, after interruption and at defined intervals
- Bead width, height, position or deposited mass against drawing requirements
- Tank, pump, line and valve pressure trends
- Vacuum level and degassing time for vacuum potting
- Valve cleaning, mixer replacement and preventive-maintenance records
- Cure time, cure temperature and hardness or functional verification
Do not force arbitrary universal numbers into the scorecard. A ±1% ratio limit may be sensible for one chemistry and wrong for another. The valid range must come from material data, qualification trials, equipment capability and the product’s functional risk.
Application scenario matrix
| Application | Material/process | Dominant failure concern | Scorecard emphasis |
|---|---|---|---|
| EV battery module | Thermally conductive 2K potting | Voids, ratio drift, incomplete fill and thermal path loss | Ratio evidence, vacuum/degassing, fill mass and cure validation |
| PCB electronics | UV adhesive, silicone or epoxy | Contamination, keep-out violation, stringing and insufficient coverage | Path accuracy, bead inspection, material handling and ESD controls |
| LED driver | Epoxy or polyurethane potting | Bubbles, incomplete cure, component stress and overflow | Degassing, fill height, cure profile and leak control |
| Automotive sensor | Gasketing or encapsulation | Seal discontinuity, adhesion loss and dimensional interference | Bead continuity, surface preparation, adhesion and traceability |
| Industrial bonding | 1K/2K adhesive | Low bond strength, open time exceeded and unstable deposit | Surface condition, mix/open time, shot mass and destructive test results |
Score containment and CAPA separately
Containment protects today’s shipment. CAPA should prevent tomorrow’s recurrence. Combining them in one vague “responsiveness” score makes it easy for a supplier to earn points for fast sorting while the real cause remains untouched.
For containment, score identification of suspect lots, stop-ship timing, inventory segregation, inspection coverage and release authority. For CAPA, score root-cause evidence, systemic causes, action ownership, due dates, implementation proof and effectiveness over representative production. A fishbone diagram is not proof. Neither is a revised work instruction if the failure involved pump wear, air ingress or an uncontrolled material condition.
A useful effectiveness test
- Can the supplier recreate or clearly explain the original failure mechanism?
- Does the corrective action remove or control that mechanism?
- Was the action tested at credible boundary conditions, not only one ideal run?
- Do records show the control working across multiple shifts and material lots?
- Could an operator or maintenance change bypass the control without detection?
Use gates as well as weighted scores
A supplier with 88 points can still be unsafe to release. Suppose delivery, documentation and response are excellent, but the supplier changed a dispense recipe without approval. The arithmetic looks acceptable; the control system does not.
| Decision state | Typical rule | Volume action |
|---|---|---|
| Restricted-red | Any red-line event or score below 70 | Pause new awards; hold or reduce volume; controlled release only |
| Restricted-amber | 70-84, no red-line event, open effectiveness evidence | Maintain capped volume and intensified inspection |
| Recovery-green | 85 or above for three periods, all gates passed | Consider phased restoration after approval |
| Normal monitoring | Stable score after phased restoration and no recurrence | Return to standard supplier surveillance |
Three periods is a practical example, but time alone is not enough. Three quiet weeks with only a handful of parts may provide less evidence than one full-rate validation run covering startup, planned stops, refill, shift change and a material-lot transition.
How long should the supplier remain restricted?
There is no honest universal answer. The duration depends on production volume, defect severity, detectability, process cycle, material shelf life and how often the relevant boundary conditions occur. The better question is: what body of evidence must exist before restrictions can change?
Define the evidence at the start. For example: three consecutive accepted lots, no recurrence of the named failure mode, completed gauge verification, successful ratio challenge, approved CAPA effectiveness review, complete genealogy records and a controlled production run at the intended rate. If a requirement cannot be demonstrated because volume is low, record the uncertainty and keep an appropriate control in place.
Who owns the scorecard?
Procurement can own the commercial status, but it should not score technical recovery alone. Quality should own defect classification, escape severity and CAPA effectiveness. Process engineering should review the dispensing window, validation and maintenance logic. Operations should confirm delivery risk and line impact. When the part affects safety, insulation, sealing or thermal management, design engineering may also need to approve the evidence.
Keep one accountable scorecard owner, even with several contributors. Otherwise the meeting ends with five spreadsheets and no decision.
Supplier review meeting checklist
- Confirm reporting period, shipped volume and denominator for every rate
- Review every defect by failure mode, severity, detection point and affected lot
- Compare actual process values with approved limits and prior periods
- Check whether sorting or rework is masking process performance
- Review overdue actions and challenge weak effectiveness evidence
- Confirm delivery, backlog, capacity and contingency stock
- Record red-line events separately from the weighted score
- Agree status, volume allocation, next evidence and named owner
- Freeze the approved scorecard version and meeting record
What buyers should avoid
First, avoid changing weights every month to match the result people want. Second, do not give full points because a document exists; check whether the control is used and effective. Third, do not let “zero complaints” stand in for process capability when shipment volume is tiny. And finally, do not restore volume as a reward for effort. Restore it when the evidence supports the production risk.
Frequently asked questions
How often should a restricted dispensing supplier be scored?
Use a weekly operational review during containment and a formal monthly scorecard. A quarterly review is too slow while defects, delivery exposure or corrective actions remain open.
What score is required before volume can be restored?
A practical starting rule is at least 85 out of 100 for three consecutive review periods, with no red-line failure. Buyers should adjust the threshold to part risk, customer requirements and validation evidence.
Can strong delivery performance offset a dispensing quality failure?
No. A critical quality escape, unauthorized recipe change or failed traceability requirement should be treated as a red-line event. Weighted totals must not hide a serious control failure.
Should PPM be the main quality metric?
PPM is useful but incomplete. Pair it with defect severity, escape location, recurrence, affected lots, containment effectiveness and process evidence such as ratio, pressure, temperature and dispense-weight records.
When should a buyer remove the restriction?
Remove it only after the supplier meets the defined score threshold, closes corrective actions with verified effectiveness, completes any required revalidation and demonstrates stable performance across representative production conditions.
What if production volume is too low to generate meaningful PPM data?
Use event-based evidence: consecutive accepted lots, representative validation runs, first-pass yield, control-chart behavior, traceability completeness and challenge tests. Do not claim stability from a tiny denominator.
Prepare the evidence before the next sourcing decision
A good scorecard makes the next action explainable. It tells the supplier what recovery means, gives engineers a common evidence set and prevents commercial pressure from quietly rewriting technical risk. If your project involves dispensing, potting, meter-mix equipment or supplier requalification, send OBO Precision the material, part drawing, current process, defect history and required output. Our engineering team can help define a practical trial and validation approach.
Related guides: when to pause or reallocate volume, when normal volume can be restored, and 30/60/90-day KPI monitoring.
Related requalification guide: use these seven recovery milestones for a restricted dispensing supplier to define evidence gates before restoring approval or production volume.
Related supplier exit guide: review when to remove a second-source dispensing supplier from the approved supplier list after recovery or integrity controls fail.
Related transfer guide: see how to transfer dispensing production after removing a supplier, including inventory, process knowledge, validation and phased release.
Related equivalence guide: learn how to prove dispensing process equivalence at a replacement supplier before releasing transferred volume.
Related inventory guide: see how to plan last-time-buy inventory during a dispensing supplier transfer.
Related customer approval guide: prepare a customer approval package for a dispensing supplier transfer before implementing or shipping changed production.
Related cutover guide: use this first-shipment cutover readiness checklist for a replacement dispensing supplier.
Next control step: After the replacement source is approved, use a documented receiving inspection plan for its first commercial shipments before unrestricted production release.
Disposition follow-up: If first-shipment inspection fails, use a controlled conditional-release decision for the replacement dispensing supplier rather than an informal use-as-is approval.
Continuity follow-up: When a failed first shipment must be rejected, use a documented production recovery plan for the replacement dispensing supplier instead of weakening the release criteria.
