Judgment · Lesson 11
Failure, mistakes, incidents, and learning
Own a real error, explain consequences and recovery, and show a durable change without blame or self-destruction.
What you will learn
- Own a real error, explain consequences and recovery, and show a durable change without blame or self-destruction.
Choose a real adverse event
A disguised success avoids the question. Select an event with an actual miss, mistaken assumption, preventable incident, or unsuccessful outcome that you can discuss safely.
- What did you get wrong?
- Who or what was affected?
- What part was not yours to own?
Learning needs a mechanism
“I learned to communicate” is incomplete. Name the test, review, alert, checklist, design constraint, or recurring behavior that changed afterward.
- What did you do immediately?
- What prevented recurrence?
- Where did the fix still fall short?
Mixed-version serialization failure
A rollout broke compatibility between old and new workers.
The team made a mistake, but I worked all night and became more detail-oriented.
I approved a schema change without testing mixed worker versions. When old workers rejected the payload, I stopped the rollout, restored the prior writer, and told support which jobs needed replay. I later added compatibility fixtures and required a mixed-version stage for schema changes.
- The personal error is explicit.
- Recovery includes users and operations.
- The lesson became a repeatable mechanism.
Write a failure timeline with decision, signal, consequence, mitigation, and prevention.
- 1Use “I” for your mistake and action.
- 2Name the affected party without exposing confidential detail.
- 3Separate immediate recovery from durable prevention.
Open the answer check
A listener should understand both the failure and why your later behavior or system is meaningfully different.
Finish this lesson
Record completion when you have worked through the material. Completion is learning progress, not evidence of mastery.
Loading completion…