IATF16949 Practical Application (22) | Customer Scorecard in the Yellow Zone: Breakdown of Scoring Criteria, Indicator Responsibility Assignment, and Red/Yellow Card Removal Criteria
At the beginning of the month, the customer portal announced the supplier score for the previous month: 78 points, in the yellow zone. The quality department printed out the scorecard and held a meeting: 6 points deducted in the quality section, 8 points in the delivery section, and 4 points in the response section. During the meeting, the quality manager said the PPM deductions were due to issues with the customer's incoming quality control (IQC) criteria, the logistics department said the delays were due to the customer's last-minute order changes, and the purchasing department said the supplier's chips hadn't arrived, and no one could conjure them. The meeting minutes filled a page, but after the meeting, no action items had a responsible person or a due date. The next month, the score was 76 points, and the following month, the company received a notice of controlled shipment from the customer, requiring 100% inspection of goods upon arrival at the customer's warehouse, with the screening costs to be borne by the company.
Such companies are often not lacking in effort, but they lack a layer of translation to convert the customer's scorecard into internal indicators. The customer's scorecard is a result indicator, and its criteria, lag, and attribution are defined according to the customer's rules. If the company directly uses it as a management indicator in meetings, the meetings will turn into sessions of mutual explanation. To implement effectively, three things are needed: a criteria mapping table, a responsibility assignment matrix, and standards for red/yellow card removal and evidence.
1. Understanding the Five Categories of Indicators and Three Scoring Mechanisms
The supplier scorecards of automotive customers may have different names (supplier performance score, supplier quality evaluation, scorecard, monthly supplier evaluation), but the indicators generally fall into five categories.
| Dimension | Common Customer Indicators | Typical Weight (Example) | Yellow Line (Example) | Red Line (Example) |
|---|---|---|---|---|
| Quality | Incoming PPM, batch acceptance rate, 0km failure count, on-site returns, warranty failure rate | 35～45 points | PPM > 50 or one batch return | PPM > 200 or safety/regulatory-related failures |
| Delivery | OTD, ASN/label accuracy, excess freight charges, emergency air shipments | 20～30 points | OTD < 95%, excess freight ≥ 2 times/quarter | OTD < 90% or causing customer production line stoppage |
| Response | Problem response time, containment action timeliness, 8D on-time closure rate, corrective action order closure period | 15～20 points | 24-hour response rate < 90% | Overdue responses ≥ 3 times/quarter |
| System & Project | IATF certificate validity, customer audit score, PPAP first-time pass rate, change notification compliance rate | 10～15 points | Customer audit < 80 points | Certificate suspension/revocation or changes without notification |
| Business & Cost | Claim deductions, rework screening cost sharing, warranty cost sharing | 5～10 points | Single-quarter claims > agreed amount | Major recall cost sharing |
The numbers in the table are just example structures. Each scorecard's threshold must be based on the customer's manual or quality agreement and clearly documented in internal files, otherwise, no one will remember where the 50 PPM threshold came from after six months.
Next, look at the scoring mechanisms. A score of 78 points can have entirely different meanings:
- Absolute Deduction System: Points are deducted from 100, with deductions tied to amounts or quantities. In this case, 78 points indicate 22 specific incidents that can be traced back.
- Ranking System: Scores are relative rankings (e.g., 22nd out of 30 suppliers). In this case, an improvement in score may not come from internal improvements but from competitors performing worse. The focus should be on "the gaps with the benchmark in which dimensions."
- Status System: Green/Yellow/Red zones, plus special status codes, such as Level 1 controlled shipment, Level 2 controlled shipment, or key monitoring. The key is the entry and exit conditions rather than the score itself.
The judgment sequence is to first determine the mechanism, then look at the details. Without a clear mechanism, focusing on the "three worst indicators" can easily lead to the wrong direction: a score drop in a ranking system may only be due to the customer raising the target value (which is a common annual practice), not because the company has done something wrong.
2. Three Key Criteria Questions to Clarify
When internal indicators do not align with customer indicators, 90% of the disputes come from three criteria.
First, what is the denominator? When the customer calculates PPM, the denominator could be the number of received items, the number of items used in production, or the number of assembled items. The unit for nonconforming items could be items, batches, or work orders. For example, in a batch of 5000 incoming items, 3 items are returned. If calculated by item, it is 600 PPM; if by batch, it is 200,000 PPM (one return per batch). Both sides are correct, but they are talking past each other.
Second, where is the lag? The score published by the customer in month M is often based on events that occurred in month M-1 or M-2, or even early nonconformities discovered in the customer's internal production line in month M. If the internal team only looks at "the current month's score corresponding to the current month's data," two consequences will arise: the effectiveness of improvement measures will be misjudged as ineffective; the customer is still processing old accounts from two months ago, while the internal team believes they have been cleared.
Third, is there any double counting? The same failure may be deducted once in the quality section as PPM, once in the warranty section as a failure rate, and once in the business section as a claim, totaling 15 points deducted, but the root cause is only one. Without consolidation, the internal team will issue three corrective action items to chase the same root cause, leading to resource waste and finger-pointing.
The solution is to create a score detail traceability table that maps each deduction in the customer's scorecard to an internally verifiable object:
| Customer Scorecard Line | Customer Description | Corresponding Customer Order/Batch | Internal Event Number | Is it Duplicated in Other Sections? | Consolidated Root Cause |
|---|---|---|---|---|---|
| Quality - Incoming Nonconformities | 12 items with seal surface scratches in March batch | Customer complaint number ××× | NC-2026-0312 | Duplicated with business section claim | Insufficient design of inner packaging |
| Delivery - Delay | Delayed by 2 days on April 8 | ASN number ××× | DL-2026-0408 | No | Customer's last-minute order change, no capacity flexibility assessment |
This table is the foundation for all subsequent work. Without it, responsibility assignment can only be based on departmental feelings, and measures will be vague, such as "improve communication."
3. Criteria Mapping: Translating Customer Indicators into Internally Controllable Indicators
The principle of mapping is that customer indicators must be traceable to an internal indicator that someone is responsible for, has system data, and has an update frequency. Customer indicators themselves should not enter internal KPIs because they are uncontrollable; internal indicators are the management objects.
| Customer Indicator | Key Customer Criteria | Corresponding Internal Indicator | Data Source | Update Frequency | Responsible Position |
|---|---|---|---|---|---|
| Incoming PPM | Number of nonconforming items in customer's incoming inspection / number of received items × 10⁶ | Outgoing batch nonconformance rate (based on customer return orders) | Customer portal + customer complaint ledger | Monthly | Customer Quality Engineer (CQE) |
| 0km Failure | Failures in customer's assembly line and before vehicle roll-off | Early failure count (within 90 days of delivery) | Customer portal + post-sales ledger | Monthly | CQE + Design/Process |
| OTD | Comparison of customer's required delivery date and actual delivery date | Commitment delivery achievement rate (based on ASN commitment date) | ERP shipping records | Weekly | Planner |
| Excess Freight | Number of times and costs of expedited freight due to supplier issues | Number of expedited shipments and costs | Logistics cost ledger | Monthly | Logistics Manager |
| Response Time | Time of first response after customer issues a problem | 24-hour response rate | Customer complaint ledger timestamps | Weekly | Customer Service/CQE |
| 8D Closure | Customer-specified closure time and quality | 8D on-time closure rate + rejection rate | 8D system | Monthly | Quality Engineer |
| Customer Audit Score | Customer process audit/system audit score | Internal audit and process audit score (same criteria) | Audit ledger | Semi-annually | System Engineer |
| PPAP First-Time Pass Rate | Proportion of first-time submissions approved | First-time submission approval rate | Project ledger | By project | Project/NPI Engineer |
| Change Notification Compliance Rate | Number of changes made without customer approval | Change notification coverage rate | Change ledger (ECN) | Monthly | R&D/Process |
Two things must be clearly documented in the mapping table, otherwise, it will become useless after six months:
- Conversion Rules (e.g., "1 batch return by the customer is converted to 300 PPM in internal indicators," which should be fixed as an internal agreement).
- Aggregation Rules (how customer return orders, internal nonconformity orders, and 8D numbers are interconnected, using a unified prefix).
4. Responsibility Assignment: Customer Deductions Should Not All Fall on the Quality Department
The most common organizational issue with customer scorecards is that all indicators are hung on the quality department: PPM is the quality department's responsibility, OTD is coordinated by the quality department, excess freight is explained by the quality department, and change notifications are pushed by the quality department. As a result, the quality department becomes an "external explanation office," with no indicator truly managed by the person who can change it.
The correct approach is to assign responsibility based on controllability and to fix it with a RACI table.
| Indicator | Responsible (R) | Approver (A) | Consultant (C) | Informed (I) |
|---|---|---|---|---|
| Incoming PPM, 0km Failure | CQE | Quality Manager | Process, R&D, Purchasing | General Manager |
| OTD, Commitment Delivery | Planner | Production/Operations Manager | Logistics, Purchasing | Sales |
| Excess Freight, Label and Packaging Violations | Logistics Manager | Operations Manager | Packaging Engineering, Planner | Finance |
| 24-Hour Response, Customer Communication | Customer Service Manager | Quality Manager | CQE, Sales | Project Team |
| 8D On-Time Closure | Department Manager | Quality Manager | CQE | Customer |
| Customer Audit Score | System Engineer | Management Representative | Process Owners | General Manager |
| PPAP First-Time Pass Rate | Project Engineer | Technical Director | Quality, Process, Purchasing | Customer |
| Change Notification Compliance Rate | R&D/Process Change Owner | Technical Director | Quality, Sales | Customer |
| Claims and Warranty Costs | Financial Cost Accountant | Financial Director | Quality, Purchasing (supplier recovery) | General Manager |
Two details that are easy to get wrong:
- Do not fill all A with the General Manager. The approver is the person who allocates resources for the indicator. If all approvals are centralized, the meeting will become a reporting session.
- The R for 8D must be the responsible department, not the quality department. The quality department writing 8D reports is the most common pseudo-implementation in the industry: if the responsible department does not write the report, the process will not truly be improved; if the quality department writes it, only the text will be changed.
5. Red/Yellow Card and Special Status Removal: Criteria and Evidence
After the customer places a supplier in the yellow zone, red zone, or controlled shipment status, the internal team often only knows they need to submit an improvement plan, but they are unclear about the criteria for removal and the form of evidence required. This must be documented in advance; otherwise, it will be a guessing game each time.
For example, the removal of controlled shipment status typically depends on four conditions (specific criteria should be based on the customer's manual):
| Removal Condition | Criteria (Example) | Required Evidence |
|---|---|---|
| Duration of Status | Continuous 3 months of scores above the green zone threshold | Three consecutive months of customer portal screenshots + internal trend chart with the same criteria |
| Root Cause Reaching System Level | Root cause is not at the individual level, such as "operator negligence" | Root cause analysis report (including evidence from 5 Why to management/design level) |
| Measure Effectiveness Verification | Comparison data before and after improvement, including sample size and observation period | Verification report: 3 batches before improvement / 5 or more batches after improvement, with stable trends |
| Cross-Checking of Similar Issues | All other customers, other production lines, and similar parts have completed the check | Cross-checking list (including evidence of "no such risk") |
| Prevention Mechanism Solidification | FMEA, control plan, work instructions, and inspection standards have been updated | Updated document list + effective date + training records |
| Confirmation of Issue Closure | Customer-side corrective action order/8D has been confirmed closed | Customer closure confirmation email or portal status screenshot |
| Customer Re-audit | On-site re-audit or document review passed | Re-audit report and nonconformity closure records |
Two practical points:
First, calculate the timeline in advance. If removal requires continuous 3 months of compliance and the improvement measures are only implemented mid-month, the actual removal point will be delayed by 4 months. If management schedules resources based on "recovery next month," they will inevitably be disappointed and start to doubt the effectiveness of the measures. The correct approach is to directly state the "expected removal month" in the improvement plan and include it as a commitment.
Second, the dates in the evidence must be consistent. The easiest way to spot inconsistencies in audits and customer re-audits is when record dates are earlier than document effective dates or verification data is later than report issuance dates. Arrange the evidence package in chronological order, with one document corresponding to one data point, to avoid "materials are present, but logic is flawed."
6. Making the Mechanism Work: 30-Minute Monthly Review Meeting
No matter how complete the system design, if it is not implemented in a fixed meeting rhythm, it will degrade into "the quality department filling out the forms" within three months. The suggested structure for the monthly review meeting is as follows, with a 30-minute duration.
| Time | Content | Input | Output |
|---|---|---|---|
| 0～5 minutes | Last month's customer score and this month's warning indicators | Customer portal data, internal indicator dashboard | Three worst indicators (list only three) |
| 5～15 minutes | Root causes and progress of the three indicators | Score detail traceability table, 8D ledger | Measures, responsible persons, due dates |
| 15～25 minutes | Review of overdue items | Measures ledger | Resource upgrade or adjustment |
| 25～30 minutes | Risk warning: potential events for deduction this month | Customer complaints, returns, and delivery anomaly list | Preventive actions and external communication criteria |
Three meeting rules are more important than the agenda:
- Discuss only controllable indicators. Customer rankings and target value adjustments are not discussion items but background information.
- Each measure must have a due date and verification method. "Strengthen incoming inspection" is not a measure; "Add 2mm EPE to inner packaging, sample 5 items for drop tests starting from the 3rd batch" is.
- Automatic upgrade for indicators not improved for two consecutive months. Upgrading is not a criticism but a trigger for resource reallocation. Write this into the system to avoid no one willing to bring it up in the meeting.
7. Indicator Decomposition: Each Function Has Its Own Quality Indicators
The true sign of a customer scorecard being implemented is not that the quality department has a beautiful dashboard, but that other departments have indicators in their evaluations that are sourced from the customer scorecard.
- The planner's KPI should include the commitment delivery achievement rate, not just the production completion rate.
- The project engineer's KPI should include the PPAP first-time pass rate, not just the on-time project launch.
- The R&D/Process KPI should include the change notification compliance rate, not just the number of changes closed.
- The purchasing KPI should include the number of nonconformities caused by suppliers, not just the procurement cost reduction rate.
At the same time, prevent two side effects:
Unidirectional indicator pressure. Raising the OTD target from 92% to 97% without providing safety stock, capacity flexibility, or the right to reject customer order changes will result in only false reporting of commitment dates to meet the target, leading to a decrease in on-time delivery from the customer's perspective. Indicator increases must be accompanied by resource allocation or rule adjustments.
Indicator conflicts. The procurement cost reduction rate naturally conflicts with the incoming PPM, and the delivery achievement rate naturally conflicts with inventory turnover. Such conflicts should be resolved at the beginning of the year using a "conflict resolution table": for example, if the annual cost reduction for the same material exceeds 8%, a supplier capability re-audit must be conducted; if emergency orders exceed 5% of capacity, the OTD evaluation should be adjusted to a weighted formula. Without a resolution rule, conflicts will ultimately be resolved through "quality concessions."
8. Key Implementation Points and Common Mistakes
| Common Mistake | Consequence | Corrective Action |
|---|---|---|
| Directly using customer scores as internal KPIs | Indicators are uncontrollable, meetings become explanation sessions | Establish a criteria mapping table, manage internally controllable indicators |
| Only looking at the current month's score | Lag leads to misjudgment, improvements are prematurely negated | Annotate the actual occurrence month of the data, aggregate by event month |
| Opening multiple corrective actions for the same failure | Resource waste, finger-pointing | Consolidate in the score detail traceability table, one root cause one project |
| Hanging all indicators on the quality department | No one truly responsible, improvements remain on paper | Assign RACI based on controllability, 8D responsibility to the responsible department |
| 8D written by the quality department | Process not changed, issues recur | Responsible department writes, quality department only reviews methods and evidence |
| Removal criteria not clearly defined in advance | Misaligned commitment times, repeated rejections | Include removal criteria and evidence list in the improvement plan |
| Measures written as "strengthen inspection/training" | Symptomatic countermeasures, not recognized in customer re-audits | Measures must address design, process, poka-yoke, or rule changes |
| Review meetings become reporting sessions | Discussion without resolution, no due dates for measures | Fixed 30-minute agenda, list only the three worst indicators |
| Indicator pressure without resource allocation | Data falsification, customer-side indicators worsen | Increase resources or adjust evaluation criteria when raising indicators |
| Changes not notified to the customer | Immediate red zone or certificate risk | Set a mandatory "customer notification" field in the change ledger (ECN), no release without notification |
The essence of the customer scorecard is that the customer scores your quality capability based on their criteria. The company can control only two things: translating that scorecard into executable internal indicators and ensuring each indicator has a person who can change it. If the translation is incorrect, efforts are wasted; if the person is wrong, meetings are futile. Red/yellow cards are never sudden; they are the result of long-term internal indicator distortions, with the customer writing the conclusion for you.
Customer scores are results; internal indicators are the levers.
Knowledge code: 2.1.2
Version: v20261009
Author: QTank QTank is dedicated to providing systematic professional knowledge, methodologies, and practical tools for quality management practitioners, helping companies continuously improve their quality capabilities.