Process Performance and Process Audit — From Lagging KPIs to True Process Method Implementation

1. Process Performance: Shifting from "Result KPIs" to "Process KPIs"

Many quality monthly reports look like this: customer complaints ↓, PPM ↓, costs ↓ — all lagging indicators (Lagging Indicators). By the time the numbers turn red, nonconforming products have already been shipped, and customer lines have already stopped.

Process performance (Process Performance) focuses on whether the process is stable, controllable, and predictable:

• Is the Cpk of critical processes up to standard?
• What is the first-pass yield (FPY)?
• What is the defect rate in the first hour after a changeover?
• What is the cycle time from ECN approval to implementation on the shop floor?

When process performance is well-managed, result KPIs often naturally improve; chasing results alone means always being in "firefighting" mode.

Process audit (Process Audit) verifies whether the process is operating as designed and whether the design is reasonable — forming a triangle with product audit (Product Audit) and system audit (System Audit).

This article provides methods for designing process performance indicators, data collection, process audit methods, and the management rhythm for linking the two.

2. Process vs. Result: Indicator Stratification

Design Principles:

1. Each end-to-end value stream (e.g., O2D, NPI) should have at least 3 process KPIs, and each should have an owner.
2. Process KPIs must be frequently updated (per shift/day/week), not just at the end of the month.
3. Each process KPI should define action thresholds (yellow/red), not just for reporting purposes.

3. Process Performance Indicator Library (Manufacturing Reference)

3.1 Production and Process

• FPY / RTY (first-pass yield/roll-through yield)
• OEE breakdown: availability, performance, quality
• SPC: Cpk/Ppk of key characteristics, out-of-control occurrences, time to correct
• Changeover time (SMED) and defect rate of the first piece after changeover
• Process parameter deviation occurrences (preferably automatically collected)

3.2 Quality Function

• NCR closure cycle time and repeat NCR rate
• Deviation / concession trends by risk level
• LPA audit compliance rate and closure rate of findings
• MSA overdue items and calibration overdue items

3.3 Supply Chain and NPI

• Incoming PPM and SCAR response cycle time
• NPI phase gate on-time rate and design change execution cycle time
• Supplier audit score trends

More indicators are not always better — 5 to 7 core process KPIs per value stream are sufficient, with the rest going into a diagnostic pool.

4. Process Performance Dashboard and Management Rhythm

It is recommended to use a three-tier dashboard:

Rules: Red items on the dashboard must have a DRI + due date; if the same reason is marked red for 3 consecutive days, it should be escalated to QRQC.

Data should be automatically collected (MES, SPC, QMS) as much as possible; manual entry should define SLA for entry and review.

5. Process Audit: Positioning and Methods

5.1 Differences from System Audit and Product Audit

The automotive industry's VDA 6.3 is a benchmark methodology for process audits; non-automotive industries can simplify and adopt its P1～P7 concepts (project management, product/process development, supplier, production, customer service, etc.), selecting processes based on risk.

5.2 Six-Step Process Audit Method

1. Select Process: Based on risk (safety, special characteristics, customer complaints, low Cpk)
2. Review Planning: Check if the control plan, FMEA, and inspection plan are consistent and up-to-date
3. Review Performance: Check SPC, FPY, and audit history trends — use data to set focus
4. On-Site Verification: Gemba observation, interviews, sample traceability (see process-specific audit walkthrough)
5. Scoring and Grading: Score each process element from 0 to 10 or grade as A/B/C nonconformity
6. Improvement and Re-audit: Measures + effect verification + re-audit cycle (based on risk)

5.3 Process Auditor Capabilities

• Understand processes and can interpret control charts and CP
• Ask questions without preconceived notions
• Distinguish between "occasional" and "systemic" nonconformities

Internally training 2～3 Lead Process Auditors is more sustainable than hiring external consultants annually.

6. Process Performance × Process Audit: Linked Closed Loop

Process performance anomaly (KPI turns red)
        ↓
  Trigger intensified process audit / special diagnosis
        ↓
  Audit findings → corrective actions → update FMEA/CP
        ↓
  Process KPI recovery → reduce audit frequency

Conversely: Process audits that find B/C class nonconformities should link to process KPIs (e.g., "LPA findings not closed"), to avoid audit reports being ignored.

It is recommended that the quality department maintain a "Process Register":

7. Relationship with Maturity and Layered Audits

• Low process maturity + red performance → highest priority for improvement and audit
• LPA verifies whether daily operations follow the CP; process audit verifies whether the CP itself is adequate — complementary, not repetitive
• The maturity model (see 2.2.3) can provide a basis for process audit frequency: Level 2 requires at least one full-element audit per year, while Level 4 can extend the cycle but intensify performance monitoring

8. Common Misconceptions

Misconception 1: Process KPIs are only for the quality department

The process owner must be the value stream responsible person (production, engineering), with the quality department providing methods and independent assessment.

Misconception 2: No audit if Cpk is good

There may be a disconnect between documents and the shop floor; good performance could also be due to overly lenient inspection standards.

Misconception 3: Copy VDA forms and fill in scores

Scoring is not the goal; improvement projects and performance recovery are.

Misconception 4: Data falsification / manual SPC modification

Audit trails and automatic data collection are defenses; data falsification should be treated as dereliction of duty.

Misconception 5: Separate meetings for audits and performance

Combine them into the value stream daily management meeting to reduce "meeting fatigue."

9. Launch Recommendations (90 Days)

10. Checklist for CQO

1. Do process KPIs appear before customer complaints/PPM in the daily report?
2. Is there a process register and is it linked to the audit plan?
3. Does the most recent process audit reference at least 6 months of SPC/FPY data?
4. Are red KPIs addressed in writing within 48 hours?
5. Are all process audit findings 100% entered into CAPA or improvement projects?

Process performance tells you "what is happening in the process"; process audit verifies "why it is happening this way and whether it should be this way". Together, they ensure the true implementation of the process approach (Process Approach) — the essence of IATF 16949.

Good process management does not wait for customer PPM alarms to call a meeting; instead, when the control chart first shows a trend deviation, the value stream team is already standing by the process.

Knowledge code: 2.2.2

Version: v20260630

Author: Quality Think Tank Quality Think Tank is dedicated to providing systematic professional knowledge, methodologies, and practical tools for quality management practitioners, helping companies continuously improve their quality capabilities.