3. Gate review fundamentals

In an APQP program, gate reviews are the formal control points that convert a plan into an auditable sequence of commitments. A gate is not a status meeting and not a progress presentation. It is a structured decision event where a cross-functional team examines objective evidence against pre-agreed criteria and produces one of a small number of authorized outcomes. The gate protects the program from two opposite failure modes: advancing with unresolved risk, and stalling on issues that do not actually block the next phase.

The purpose of a gate can be summarized in four functions. First, verification of phase deliverables: confirming that the outputs required by the APQP phase plan exist, are complete, and meet the defined quality standard. Second, risk disposition: deciding whether open issues are acceptable to carry forward, must be closed before exit, or require escalation. Third, resource and schedule authorization: releasing budget, tooling, headcount, or supplier commitments for the next phase. Fourth, traceability: producing a signed record that links requirements, design decisions, verification evidence, and approvals. Without this record, later problem-solving during launch or warranty has no reliable baseline.

Timing: Gates Follow Evidence, Not Calendars

Gate timing is anchored to phase completion, not to a fixed date on a project schedule. In a typical APQP structure aligned with AIAG-style phasing, gates are placed at the boundaries between:

• Plan and Define → Product Design and Development
• Product Design and Development → Process Design and Development
• Process Design and Development → Product and Process Validation
• Product and Process Validation → Feedback, Assessment, and Corrective Action
• Launch readiness → Production

A gate may be scheduled on a calendar for planning purposes, but it should only be convened when the entry criteria are demonstrably met. If the team holds a gate before evidence exists, the review degenerates into a promise-making session. If the team delays indefinitely waiting for perfect evidence, the program loses its control function. The practical rule is: schedule the gate, but let the entry criteria determine whether it opens as a decision review or converts into a readiness assessment with a defined closure plan.

Entry and Exit Criteria

Entry criteria define what must be true before the review can begin. Exit criteria define what must be true before the phase can be formally closed and the next phase authorized. These are different from deliverables. A deliverable is an artifact; a criterion is a condition on that artifact or on the program state.

Consider a Design Gate between product design and process design. Representative entry criteria might include: design FMEA completed to a defined severity and occurrence threshold; design verification plan signed; critical characteristics identified and flowed into drawings; prototype build completed with dimensional reports available. Exit criteria might include: all severity 9–10 failure modes have either a verified design control or an approved deviation; design verification test results meet specification with no open non-conformances above the agreed threshold; process design team has received the finalized characteristic matrix; and the gate decision record is signed by the accountable roles.

The distinction matters because teams often treat "the document exists" as sufficient. A DFMEA that exists but has no action closure for high-severity items does not satisfy an exit criterion. A dimensional report that exists but shows out-of-specification results without disposition does not satisfy an exit criterion either.

Decision-Making Roles

Gate decisions require clearly assigned authority. A common and workable model separates three role types:

The Gate Owner must have the authority to say no, or the gate is decorative. In many organizations the Gate Owner is also accountable for escalation when a criterion cannot be met and the program must proceed under a documented deviation. That deviation must name the risk, the compensating control, the owner, and the expiry condition.

Decision Outcomes

A gate should produce one of a defined set of outcomes, not a vague "continue with actions." A practical set is:

• Pass: all exit criteria met; next phase authorized.
• Conditional Pass: a bounded set of non-critical items remain open with named owners, due dates, and a defined verification method; next phase authorized with conditions.
• Hold: one or more critical criteria not met; next phase not authorized until a re-review confirms closure.
• Redirect: evidence reveals a design or process issue that requires returning to an earlier phase activity before the gate can be reconsidered.

The difference between Conditional Pass and Hold is not severity of the issue alone; it is whether the open item can be closed in parallel with the next phase without invalidating work already done or planned. If parallel closure would create rework in the next phase, the correct outcome is Hold.

Making the Gate Operable: A Worked Example

Suppose a Tier 1 supplier is preparing a Process Design Gate for a stamped and welded seat frame. The entry criteria require: process flow diagram released, PFMEA completed with all severity ≥ 8 items having recommended actions assigned, control plan draft aligned to PFMEA, and gauge R&R studies completed for all critical characteristics. The exit criteria require: all severity ≥ 8 PFMEA actions closed with verification evidence, control plan approved by manufacturing and quality, gauge R&R results below 10% for critical characteristics or an approved containment plan, and operator training matrix updated.

During the review, the team finds that gauge R&R for the weld nugget diameter measurement shows 18% repeatability and reproducibility. The PFMEA action for weld parameter monitoring is closed, but the measurement system is not capable. The Gate Owner cannot issue a Pass because the exit criterion is not met. The team evaluates whether this can be a Conditional Pass. Because the next phase includes process validation builds that depend on this measurement for acceptance, parallel closure would invalidate the validation data. The correct outcome is Hold, with a defined action: complete a gauge improvement study, re-run R&R, and reconvene the gate within two weeks. The recorder documents the decision, the action owner, the due date, and the re-review date.

This example illustrates the core discipline: the gate decision follows from evidence against criteria, not from schedule pressure or optimism. The Gate Owner's authority is meaningful only when the criteria are specific enough to make the decision defensible.

Common Failure Patterns

• Criteria written as activities: "Complete DFMEA" instead of "All severity 9–10 failure modes have verified controls or approved deviations." The first cannot be failed objectively.
• Gate Owner without authority: the decision is escalated to a steering committee, and the gate becomes a reporting layer rather than a control point.
• No record of conditions: Conditional Pass items are discussed but not tracked, so they reappear at the next gate as unresolved risk.
• Gate held before entry criteria are met: the review becomes a planning meeting, and the actual decision is deferred without a formal re-review date.

Well-designed gates do not slow a program down. They prevent the far more expensive pattern of discovering unresolved risk during tooling, validation, or launch, when the cost of correction is an order of magnitude higher. The gate is where the program earns the right to proceed.