QG — Quality Gate

In plain English

Quality Gate (QG) is a checkpoint where work must meet clear quality rules before it can move to the next step. It exists to stop defects early, when they are cheaper and easier to fix. A QG defines what “good enough” means, who checks it, and what evidence is required.

It works by setting pass/fail criteria, doing a review or test, and recording the result. If it passes, the work continues. If it fails, the work stops and gets corrected. A good QG also defines what happens when something is unclear, including who can make an exception and how that exception is documented. The goal is consistent output, not blame.

What they actually mean

On paper, a QG prevents bad work from moving downstream.

In reality, it often turns into a calendar meeting that “approves” whatever is in front of it because the schedule is already committed.

• Criteria are vague (“looks good”) or change mid-stream
• Evidence is a slide deck, not test data or inspection records
• People show up without decision rights, then “take it offline”
• Failures get re-labeled as “risks” so the gate can still pass

Uncomfortable truth: If failing the gate is not allowed, it is not a gate. It is theater.

Another common failure is stacking QGs to compensate for weak process control. That creates delays, more handoffs, and more chances to lose the actual facts.

When done right, the QG is boring and fast: clear pass/fail criteria, objective evidence, one accountable owner, and a defined containment path when it fails. The gate protects the customer and the team because it makes “stop and fix” normal.

Example

Where you’ll hear it

Used in product development, industrialization, manufacturing release, software delivery pipelines, supplier qualification, and change control—anywhere a handoff can hide defects.

“We don’t move to the next phase until the QG evidence is in the folder and signed off.”

Does it actually matter?

✅ Yes — when the work is high-risk, expensive to rework, or hard to inspect later (safety, regulatory, customer-critical performance).

Quality Gates matter because they force explicit criteria, objective evidence, and a decision point before you multiply defects at scale. They also clarify ownership: who decides, based on what, and what happens if it fails.

⚠️ Watch out: If leadership won’t tolerate a fail, or the criteria are subjective, QGs become delay plus paperwork. Then people learn to game the gate instead of improving the process.

If you're working with quality gates but lack structured evaluation tools, this book gives you the practical methods behind the checkpoints.
Less theory. More “how to actually decide.The Quality ToolboxThis book provides tools that are less commonly used and some tools that the author, Nancy Tague, created. Inside you'll find tools for generating and organizing ideas, evaluating ideas, analyzing processes, determining root causes, planning, basic data hRecommended (affiliate)

Common misconceptions

• Myth: A QG is a meeting.
  Reality: It is a decision based on defined criteria and evidence; the meeting is optional.


• Myth: Passing means “no issues.”
  Reality: Passing means criteria were met; known risks must be documented with owners and dates.


• Myth: More gates means more quality.
  Reality: Too many gates often means weak process control and slow learning loops.


• Myth: A senior leader’s approval is the evidence.
  Reality: Evidence is test data, inspection records, and documented requirements.


• Myth: Failing the gate is a performance problem.
  Reality: Failing is a normal signal to contain and correct before scaling.

Red flags

• 🚩 Criteria like “reviewed” or “looks good.”
  Problem because pass/fail becomes opinions, and defects slip through on confidence alone.


• 🚩 Exceptions are routine and undocumented.
  Problem because the gate stops being a control point and turns into a habit of unmanaged risk.


• 🚩 No defined owner with decision rights.
  Problem because issues bounce between functions, and the schedule makes the decision by default.


• 🚩 Evidence is slides, not records.
  Problem because you can’t audit it, repeat it, or learn from it when something fails in the field.


• 🚩 Gate failures don’t trigger containment.
  Problem because the same defect pattern repeats downstream, just with higher cost and more customers involved.

Worth learning?

4/5

Worth learning because it’s a practical control method for handoffs and scaling. It works best when criteria are measurable and failing the gate is treated as normal process feedback.