Chapter 31: Designing a Verification Plan Before You Close the CAR

This is the operational heart of effectiveness verification, and it's where many teams fail because they treat it as an afterthought. The verification plan is not something you write after the fix is in place. You write it *as part of the root cause analysis*. It lives in your CAR documentation alongside the RCA and the corrective action description.

A complete verification plan contains these non-negotiable elements:

The Metric to Be Measured

This is the specific, measurable indicator that proves the root cause is gone. Not "product quality" or "process stability." Specific. If your root cause was tool wear causing dimensional drift, your metric is the specific dimension (e.g., 23.450 ± 0.050 mm). If your root cause was operator technique on an assembly step, your metric might be "torque value at final tightening" or "gap measurement at joint interface" or even "defect escape rate on next 1,000 units."

The Target Value or Acceptance Criteria

What does "gone" look like numerically? If you're monitoring an SPC chart (Xbar-R, individuals, or p-chart), your target might be "zero points outside control limits over 60 days." If you're monitoring capability, your target might be "Cpk ≥ 1.33 sustained." If you're monitoring attribute data, your target might be "zero defects on sample size of 125 units per shipment for 90 days."

The Measurement Method

How will you collect the data? Will you measure every piece? Every 50th piece? Every first piece and last piece of each production run? Will you use a specific gauge, CMM, or test equipment? Will you use historical data, or will you collect new data? Document this precisely. Vagueness here is where auditors find gaps.

Responsible Person and Verification Deadline

Someone owns this. Not "the quality team." A specific name, title, and email. And a specific deadline: "Verification complete by June 15, 2026" beats "ongoing monitoring" because it creates accountability and a checkpoint.

Pro Tip: A Windsor, Ontario automotive supplier fixed a nonconformance tied to insufficient training on a new assembly fixture. Instead of vague "ongoing monitoring," they specified: FPY ≥ 98% sustained over 30 consecutive production days, 100% dimensional verification of first three pieces daily, responsible person Sarah Chen (Quality Engineer), verification deadline November 15, 2026, with daily SPC charts and monthly statistical capability analysis. That specificity is what turns a CAR from ambiguous to auditable.

Let's ground this further. A major OEM supplier in Windsor had a root cause: insufficient training on a new assembly fixture. Here's how they designed their verification plan:

• Metric: First-pass yield (FPY) on Assembly Process 4B using the new fixture
• Target: FPY ≥ 98% sustained over 30 consecutive production days
• Measurement Method: 100% dimensional verification of first three pieces off the fixture each day; defect tracking via the existing MES system
• Sample Size: Minimum 120 assemblies measured across the 30 days
• Responsible Person: Sarah Chen, Quality Engineer, sarah.chen@[company].ca
• Verification Deadline: November 15, 2026
• Documentation: Daily SPC chart (individuals chart) plotted with control limits; monthly summary report with statistical capability analysis

That's a verification plan. It's tight, measurable, and auditable.

Important: Statistical methods aren't optional for manufacturing effectiveness verification. They are the language auditors speak and the standard your certification body expects. Familiarize your team with Xbar-R charts (subgroup data), individuals charts (X-MR for single-piece monitoring), p-charts (proportion defects), and capability indices (Cpk, Ppk). These tools transform "things are better" into "things are statistically better." For automotive suppliers, the AIAG statistical handbook and IATF 16949 guidance provide the reference standard your auditor will use to evaluate your evidence.