Most companies evaluate PCB suppliers the wrong way. They compare price per board, check if the delivery was on time, and call it supplier management. When a quality problem surfaces — and it always does — they have no quantitative basis for deciding whether this is an anomaly or a pattern. The result: suppliers are fired too late, kept too long, or replaced for the wrong reasons.
A supplier quality scorecard solves this. It converts the four dimensions of supplier performance — quality, delivery, cost, and responsiveness — into weighted scores that trend over time. A single bad month doesn't trigger a panic. A slow decline across three quarters is visible six months before it becomes a crisis. This article provides the framework, formulas, and thresholds we use at Huaxing PCBA — refined from evaluating our own material suppliers and from the scorecards our largest automotive and medical customers use to evaluate us.
Why Price-Per-Board Is an Incomplete Metric for PCB Sourcing
The unit price of a PCB captures approximately 60% of its total cost of ownership. The remaining 40% is hidden in quality failures, late deliveries, rework, line-down costs, and procurement overhead. A supplier quoting $8.50 per board with a 2% defect rate actually costs $9.35 per good board when factoring in rework and scrap. A supplier at $9.00 with a 0.5% defect rate costs $9.05 per good board — cheaper in practice despite the higher unit price.
The cost of quality formula: True cost per board = Unit Price ÷ (1 − Defect Rate) + (Rework Cost × Defect Rate). For a $10 board with 3% defects and $5 rework cost: $10 ÷ 0.97 + $5 × 0.03 = $10.31 + $0.15 = $10.46. That $0.46 difference looks small — until you multiply it by 50,000 boards per year. Then it is $23,000 of hidden cost that never appears on any invoice.
This is why a proper scorecard weights quality above cost. If your scorecard doesn't, you're optimizing the wrong variable. For a complete cost analysis framework, see our PCB cost factors guide.
The 4-Pillar Scorecard Framework
The scorecard uses four weighted pillars, each scored 0-100 and combined into a single composite score. The weightings below are our recommended defaults for a PCB/PCBA supplier; adjust them based on your product's risk profile.
| Pillar | Weight | What It Measures | Data Source |
|---|---|---|---|
| Quality | 40% | Defect rate, first-pass yield, field returns, CAPA effectiveness | IQC records, production data, RMA logs |
| Delivery | 25% | On-time delivery, lead time accuracy, quantity accuracy | ERP purchase order data |
| Cost | 20% | Price trend, cost reduction initiatives, NRE transparency | Quote history, invoice data |
| Responsiveness | 15% | Quote turnaround, problem resolution speed, engineering support quality | Internal tracking, email response metrics |
Composite Score = (Quality × 0.40) + (Delivery × 0.25) + (Cost × 0.20) + (Responsiveness × 0.15)
The result is a single number between 0 and 100. Track it quarterly. Trend it over 4+ quarters. A supplier dropping from 85 to 78 over two quarters is a conversation. A supplier dropping from 85 to 65 in one quarter is an escalation.
Pillar 1: Quality (Weight: 40%)
Quality is the highest-weighted pillar because quality failures in PCB supply have the longest tail of consequences. A late board delays a shipment. A defective board that passes incoming inspection and fails in the field costs 10-100× the board price in warranty claims, reputation damage, and — for regulated industries — regulatory exposure.
Incoming Lot Acceptance Rate (ILAR) — 40% of Quality Score
ILAR = (Accepted Lots ÷ Total Received Lots) × 100
Score: ≥99% = 100 | 98-98.9% = 85 | 97-97.9% = 70 | 95-96.9% = 50 | <95% = 25
This is the bluntest instrument in the quality toolkit — and the most important. A supplier whose lots are rejected 3% of the time is creating re-inspection, MRB meetings, NCMRs, and supplier corrective action requests on 3% of everything they ship. Our incoming quality inspection guide covers sampling plans and acceptance criteria.
Defective Parts Per Million (DPPM) — 30% of Quality Score
DPPM = (Defective Units Found ÷ Total Units Inspected) × 1,000,000
Score: <500 DPPM = 100 | 500-1,000 = 85 | 1,001-2,500 = 70 | 2,501-5,000 = 50 | >5,000 = 25
DPPM captures defects within accepted lots — the subtle quality erosion that ILAR misses. A supplier shipping lots that pass AQL sampling but have an underlying DPPM trend rising from 800 to 1,200 over three quarters is in trouble, even if every lot passes incoming inspection.
Corrective Action Effectiveness — 20% of Quality Score
Score: All CAPAs closed on time with verified effectiveness = 100 | >90% closed on time = 80 | 75-89% closed on time = 60 | <75% closed = 30
This measures whether the supplier actually fixes problems — not whether they write a nice 8D report. The test: when you reject a lot for a specific defect, does that defect recur in the next three shipments? If yes, the CAPA was paperwork, not process improvement. Our PCB failure analysis guide covers root cause methodologies.
Field Return Rate — 10% of Quality Score
Field Return Rate = (Field Returns ÷ Total Shipped Units) × 100 (12-month rolling)
Score: <0.1% = 100 | 0.1-0.3% = 80 | 0.31-0.5% = 60 | 0.51-1.0% = 40 | >1.0% = 20
The lagging indicator. A quality problem that escapes IQC and reaches the field is a system failure — not just a supplier failure. For PCB failure modes that reach the field, see our PCB rework and repair guide.
Pillar 2: Delivery (Weight: 25%)
Delivery performance in PCB supply is not binary (on time / late). A shipment that arrives on the right date with the wrong quantity is a partial failure. A shipment that arrives early — filling your warehouse before you're ready — is also a failure. The scorecard must capture nuance.
On-Time Delivery (OTD) — 50% of Delivery Score
OTD = (Line Items Delivered On Time ÷ Total Line Items) × 100
Score: ≥98% = 100 | 95-97.9% = 85 | 90-94.9% = 70 | 85-89.9% = 50 | <85% = 25
"On time" means within the agreed delivery window — typically ±1 business day for domestic shipments, ±2 days for international. Early delivery counts as on-time only if pre-approved. A surprise early shipment that arrives before your receiving team is staffed is a nuisance, not a favor.
Quantity Accuracy — 30% of Delivery Score
Quantity Accuracy = (Line Items with Correct Quantity ÷ Total Line Items) × 100
Score: ≥99% = 100 | 97-98.9% = 85 | 95-96.9% = 70 | 90-94.9% = 50 | <90% = 25
PCB suppliers commonly ship ±10% of ordered quantity — the industry convention is that overage/underage within this band is acceptable. But if your MRP system doesn't handle variable receipt quantities well, a 5% shortfall can cause a line-down situation. Negotiate quantity tolerance with your supplier and bake it into the scorecard.
Lead Time Stability — 20% of Delivery Score
Measured as standard deviation of actual lead time over 6 rolling orders
Score: σ < 1 day = 100 | σ 1-2 days = 85 | σ 2-3 days = 70 | σ 3-5 days = 50 | σ >5 days = 25
A supplier with a stable 10-day lead time is more valuable than one with an average 7-day lead time that varies between 4 and 18 days. Predictability enables your production planning. Average lead time that you cannot plan around creates buffer inventory costs that more than offset the faster nominal lead time.
Pillar 3: Cost (Weight: 20%)
Cost scoring is a trap if done wrong. The most common error: scoring the lowest-price supplier at 100 and everyone else proportionally lower. This drives supplier selection toward the cheapest option — which is exactly what the quality and delivery pillars are there to prevent.
Correct approach: Score cost on value trend, not absolute price. A supplier whose unit price increases 2% annually with documented material cost justification scores higher than one whose price stays flat but whose quality is deteriorating. The cheapest supplier gets a 70, not a 100, unless they also lead on quality and delivery. The scorecard is designed to reward balanced performance, not cost minimization.
Price Competitiveness — 40% of Cost Score
Benchmark against 3 comparable quotes for the same specification. Score: within 5% of median = 100 | within 5-10% above = 80 | 10-15% above = 60 | >15% above = 40. Being >10% below median is also scored at 80 — abnormally low pricing signals corner-cutting or unsustainably thin margins. Our guide to PCB price negotiation covers quote benchmarking in detail.
Cost Reduction Initiative — 30% of Cost Score
Has the supplier proposed ≥2 cost reduction ideas in the past 12 months? Score: 2+ implemented ideas saving ≥3% = 100 | 1-2 proposed = 70 | none = 40
A supplier who proactively suggests panelization optimization, material substitution, or test strategy changes that reduce your total cost is worth more than a supplier who simply holds price. The best PCB suppliers think about your cost structure, not just their invoice. Our PCB panelization cost optimization guide is one example of what a good supplier should be proposing.
NRE and Quote Transparency — 30% of Cost Score
Score: NRE charges itemized with justification = 100 | lump sum NRE with verbal explanation = 70 | NRE appears to include hidden margin = 40
Tooling, test fixtures, and engineering charges are legitimate costs. But when NRE is a lump sum with no breakdown, it is often a margin opportunity for the supplier. A transparent quote separates one-time costs from recurring costs — and explains why each NRE line item exists. For a deep dive on NRE structure, see PCB NRE and tooling costs.
Pillar 4: Responsiveness (Weight: 15%)
Responsiveness is the lowest-weighted pillar — not because it's unimportant, but because it is the most subjective. Raw response speed without quality is dangerous; a supplier who answers every email in 10 minutes but ships defective boards is worse than one who responds in 4 hours and ships good product. Weight responsiveness less than quality, delivery, and cost.
| Metric | Weight in Pillar | 100 Points | 50 Points | 0 Points |
|---|---|---|---|---|
| Quote turnaround (RFQ to formal quote) | 35% | <24 hours | 24-48 hours | >72 hours |
| Quality issue response (acknowledgment + containment plan) | 35% | <24 hours | 24-48 hours | >72 hours |
| Engineering support (DFM feedback on new designs) | 20% | <2 business days, with specific recommendations | 2-4 days, general comments | >1 week or no feedback |
| Proactive communication (schedule changes, material shortages) | 10% | Notified before you discover the problem | Notified after you ask | Never notified |
Scorecard Thresholds: When to Act
The composite score is not the end product — the trend is. A supplier at 82 and flat for four quarters is healthy. A supplier at 85 but declining 3 points per quarter is a problem you need to address now, not when they hit 70.
| Score Range | Classification | Action |
|---|---|---|
| 90-100 | Strategic Partner | Increase volume share, involve in NPI, negotiate long-term agreement. These suppliers are rare — protect the relationship. |
| 80-89 | Preferred Supplier | Maintain volume, conduct annual business review, identify one joint improvement initiative per year. |
| 70-79 | Approved — Watch | Cap volume at current level, increase incoming inspection sampling, require monthly quality data submission. Flag for quarterly review. |
| 60-69 | Conditional — Improvement Required | Formal improvement plan with 90-day milestones. Reduce order volume 30-50%. Begin qualifying alternative supplier. Our supplier transition guide covers the qualification process. |
| <60 | Disqualification Candidate | Stop new PO placement. Complete outstanding orders under heightened inspection (AQL tightened by one level). Execute transition to qualified alternative. See our dual-sourcing strategy for transition planning. |
The one-quarter rule: A scorecard is a 4-quarter trailing tool, not a single-quarter snapshot. Never disqualify a supplier based on one bad quarter — a single major customer order going wrong can tank quality metrics temporarily. But also never let a supplier explain away three consecutive declining quarters as "temporary." Three quarters is a trend. A trend requires a corrective action plan with defined milestones. If the milestones are missed, the fourth quarter triggers escalation.
Implementation: Start Simple, Add Sophistication Over Time
Do not build a perfect scorecard before you start using one. The version 1.0 scorecard needs only: lot acceptance rate, on-time delivery, and price trend. These three metrics — tracked monthly, reviewed quarterly — will surface 80% of supplier performance issues. The remaining 20% (DPPM, CAPA effectiveness, lead time stability) can be added as your data collection matures.
The most common implementation failure: building a beautiful scorecard in Excel, populating it with one quarter of data, and never updating it. A scorecard updated once and abandoned is worse than no scorecard at all — it creates the illusion of supplier management without the substance. Schedule a recurring calendar event. Assign ownership to one person. Review at quarterly business reviews with the supplier present. Our supplier audit checklist provides the on-site verification that scorecard data alone cannot deliver.
At Huaxing PCBA, we apply this framework internally — to our material suppliers — and we are evaluated against it by our largest automotive and medical customers. A supplier who understands how they are being scored is a supplier who can improve. A supplier who has no idea how you evaluate them is a supplier you haven't really managed.
