Change Review is Only the First Half of the Game — Five Steps for Change Implementation and Closure

A company recently felt aggrieved: the change review was conducted diligently—CCB meetings, impact analysis, and signatures on every item, without a single step missed. After the ECN was approved, the technician worked overnight to modify the program and replace the tooling. Two trial pieces were dimensionally qualified, and the next day, the batch production was notified. However, a week later, the entire batch was returned by the customer due to assembly interference. Upon review, it was discovered that the two trial pieces were made from old batch blanks, and the dimensional deviation of the new blanks' positioning surfaces was just "offset" by the old program; after the batch switch, the program remained unchanged, and the interference immediately became apparent.

The change review ensures that the change is necessary, but no one checks if the change is effective after implementation. Too many companies focus their change management efforts on the review, assuming "approval means success," thus the high-risk area for change failures quietly shifts to the post-approval phase: verification, trial operation, switching, and closure. This article outlines a five-step method to ensure these four stages are properly managed.

1. After Approval, the Risk of Change Failure is High

The change review answers "can it be changed," while implementation verification answers "is the change reliable." Both are essential, but the latter is often reduced to a simple "first article inspection."

Failures at the site usually occur due to three broken links. First, the verification link: only verifying the product, not the process. If the dimensions are qualified, the process is released, but the process capability, inspection tools, poka-yoke, and cycle time are not confirmed, leading to issues when the batch production starts. Second, the scope link: only verifying the specific design change, not the associated items affected—changing the supplier's material without verifying its machinability; changing the hole diameter without validating the welding fixture and inspection tools. Third, the closure link: verification is done, but results are not recorded, reviewed, or closed, and after three months, no one can clearly state what was verified initially.

Passing the review is just the "starting signal." Verification and closure are the full-process supervision that ensures the change can safely reach its destination.

2. Step One: Translate "What Has Changed" into "What Needs to Be Verified"

Verification should avoid "making decisions based on intuition." The correct starting point is the change impact analysis table—items affected have already been identified during the review, and these should be translated into verification tasks during implementation.

The translation should cover three categories. First, product verification: what are the changed dimensions, functions, performance, and reliability metrics, and what corresponding tests or measurement items and criteria (drawing tolerances, test standards, customer specifications) are required. Second, process verification: which process parameters, tooling, programs, inspection tools, and poka-yoke devices have changed, and whether capability studies, GR&R, and first article inspections are needed. Third, document and system verification: whether FMEA, control plans, work instructions, and inspection specifications have been updated, and whether the site is using the latest or old versions.

Each item should specify "who, what, by what standard, and what evidence" to form a verification task list. A special reminder: verification must be conducted under formal production conditions (final tooling, program, formal blanks) rather than "temporary conditions"; sample pieces should not be single pieces—single-piece qualification does not prove batch stability; at least one minimum verification batch should be covered.

3. Step Two: Trial Operation Verification—Prove "the Process Can Hold Up" Under Production Conditions

Product verification addresses "whether the product is correct," while trial operation addresses "whether the process is stable." Any change involving process, tooling, program, material, or fixture changes should be followed by a trial batch, which should be run under formal production conditions (people, machines, materials, methods, environment, and measurement).

During the trial operation, four key points should be confirmed. First, first article and full-size inspection: the first piece in the new state must undergo full-size and full-performance inspection, with measurement records retained. Second, process capability: short-term capability studies should be conducted for key and special characteristics, typically requiring Cpk≥1.33, and special characteristics should meet the customer's requirement of 1.67; the measurement system must pass MSA to ensure data reliability. Third, poka-yoke and fixture effectiveness: if the change involves poka-yoke devices or fixture positioning, "destructive verification" should be actively performed—intentionally using wrong parts or positions to test if the poka-yoke fails. Fourth, cycle time and operational feasibility: if the process changes, whether the operator's work sequence, cycle time, and safety are affected, and whether the new work instructions have been adequately trained.

Trial operation is not just a "formality to produce a batch," but a run with a problem list: issues discovered should be recorded and addressed step-by-step, and any issues affecting release must be rectified and re-verified; products from the trial batch can only be released if they are qualified, and any suspicious items must be isolated and reviewed.

4. Step Three: Product Verification and Customer Approval—Present Evidence Externally

Internal verification is only half the battle. Changes involving customer requirements, regulatory requirements, or product interfaces must also complete external approval, which is the easiest step to overlook.

External verification has three layers. First, conduct all tests according to customer specifications: in addition to dimensions, functional, durability, environmental, and material tests must not be skipped, and they must cover the formal state after the change. Second, submit for approval according to the agreement: in the automotive industry, PPAP submission is typically required, submitting samples, test reports, capability studies, control plans, PSWs, etc., according to the customer's required level; even if the customer only requires a "notification level," complete evidence should be retained for review. Third, identify customer special requirements (CSR): some customers have specific timelines, formats, and "no mass production before approval" red lines; missing any of these can result in batch non-compliance.

Don't forget the partners in the chain: if the change affects sub-suppliers, their tooling, processes, and inspection tools must also be verified and evidence obtained. External evidence should be archived in the change record and used as one of the inputs for closure—switching to the new state without customer approval is a common high-risk item in audits and customer complaints, and this red line must be strictly enforced.

5. Step Four: Breakpoint Switching and Initial Management—Contain Risks in the "Isolation Zone"

Verification must be fully completed before switching. Switching is not as simple as "notifying the workshop to use the new state tomorrow," but a managed "track change" within the manufacturing system.

Breakpoint management must address three questions. First, where does the old state stop: clearly define the breakpoint number or batch, and how to handle the remaining old state materials, work-in-progress, and finished products—whether to use them up, isolate them, or rework them, a written decision must be made to prevent mixing of old and new states. Second, where does the new state start: the switching time, first batch identification, and first article inspection responsibilities must be specified; the "recovery of old versions and distribution of new versions" of inspection tools, programs, and work instructions must be completed and recorded. Third, whether the supplier is synchronized: if the sub-supplier's breakpoint is inconsistent with the factory, incoming materials may have "new drawings with old parts," and incoming inspection should be marked according to the breakpoint state.

Initial management after switching is about containing residual risks in the isolation zone: a clear tracking period (e.g., one week to one month) should be arranged, with stricter controls during this period—full inspection or increased sampling, enhanced patrol inspections, daily tracking of defect rates and anomalies; any signs of issues should be escalated quickly, and small problems should be addressed on the same day to prevent batch customer complaints. Products during the tracking period should be separately marked and traced until stability is confirmed, then they can return to normal control.

6. Step Five: Evidence Closure and Closure—Ensure Each Change Has a "Record"

No matter how comprehensive the verification, it is meaningless without closure. Closure is the "final inspection" of change management: verify each item against the initial verification task list, supplement any missing evidence, and redo any nonconformities. Once all requirements are met, the CCB or authorized person should conduct the closure review and formally announce "change closed, transition to normal production control."

Three things should be done simultaneously during closure. First, update the history: record the breakpoint number, switching date, verification conclusion, and closure date in the ECN ledger to ensure each product can answer "which change state was used." Second, confirm document closure: ensure that FMEA, control plans, work instructions, and inspection specifications are consistent with the post-change state, and all old documents are recovered and invalidated. Third, share experiences: synchronize the lessons learned and effective verification methods with similar products, production lines, and suppliers to turn one lesson into a broader defense; companies with the resources should also arrange a review of tracking period indicators to seal the "closure" with data.

7. Don't Let Verification Become a "Performance": Four Common Misconceptions

Misconception One: Using trial cutting as a trial operation. A few handmade pieces do not prove batch stability; trial operation must be conducted in batches under production conditions. Misconception Two: Verifying only "the changed part." Changing the hole diameter affects the welding fixture, inspection tools, and supplier, so the verification scope should follow the impact analysis. Misconception Three: Supplementing customer approval after the fact. Switching without prior approval can lead to batch scrapping and reputational damage, the cost of which far exceeds waiting. Misconception Four: Closing without complete evidence. Verbal assurances of "no problem" are not sufficient; closure must be based on traceable records.

The review decides whether the change can "start," while verification decides whether the change can "arrive." Completing these five steps—verification plan, trial operation, customer approval, breakpoint switching, and evidence closure—ensures that the change truly lands as an improvement, not an accident.

The review manages the start, and verification manages the arrival—completing the five steps of the verification plan, trial operation, customer approval, breakpoint switching, and evidence closure ensures the change truly lands.

Knowledge code: 8.4.1

Version: v20260903

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.