Why does the same defect keep coming back after we "fixed" it?

Here is a pattern that shows up across aerospace, automotive, and electronics plants. A defect appears. A team opens a corrective action, contains it, finds a cause, writes a fix, and closes it. Weeks later the same defect returns, and everyone concludes the fix failed.

It usually did not. The fix addressed the cause the team found, not the cause that mattered.

The 5-Why that stops at three

Root cause analysis is uncomfortable by design. The honest version keeps asking after the answer is convenient. Most investigations stop at the third why, because that is where the blameless-enough answer lives, such as the operator missed it or the gage drifted. Those are symptoms wearing a cause's clothing.

Three tests for a real root cause

• If your corrective action is retrain the operator, you have almost certainly stopped too early.
• A true root cause, when removed, makes the defect impossible, not just less likely.
• The fix is verified with data months later, not closed on a signature the same week.

Closing the loop

The difference between closing a problem and closing the loop is verification that holds over time. Containment stops the bleeding. Root cause removes the wound. Verification proves it stayed closed. Skip the third step and you will meet the same defect again, usually in front of a customer.