Skip to main content

> Term

5 Whys

An iterative causal interrogation technique that explores cause-and-effect relationships by repeatedly asking "Why?", structured with branching rules and prohibiting human error as a terminal cause.

Detailed Explanation

The 5 Whys is an iterative interrogative technique developed within the Toyota Production System and adapted by Site Reliability Engineering to peel back layers of symptoms leading to systemic root causes. By asking "Why did this happen?" repeatedly, investigators traverse the causal chain from visible defect to architectural vulnerability.

In production software RCA, the 5 Whys enforces three non-negotiable rules: 1) Every answer must be substantiated by timestamped telemetry. 2) Attributing the terminal why to human error ("engineer forgot flag") is strictly forbidden. 3) If an answer involves multiple concurrent conditions, the interrogation must branch into parallel causal paths.

Why It Matters

Forces engineering teams to look past surface symptoms and address the underlying organizational, architectural, or automation failures.

Common Failure Mode

Stopping at the first or second "Why" (settling for a proximate cause), or using linear interrogation when failure required multiple converging conditions.

Practical Example

Drilling down from "API returned 504" to "Database pool exhausted" to "Connection leak in retry logic" to "Missing automated leak test in CI/CD pipeline".

Production Manifestation

Hierarchical 5-step causal trees embedded in postmortem documents and RCA engineering reviews.

Frequently Asked Questions

What is 5 Whys in short?

An iterative causal interrogation technique that explores cause-and-effect relationships by repeatedly asking "Why?", structured with branching rules and prohibiting human error as a terminal cause.

What is the most common failure mode?

Stopping at the first or second "Why" (settling for a proximate cause), or using linear interrogation when failure required multiple converging conditions.

AI Summary

An iterative causal interrogation technique that explores cause-and-effect relationships by repeatedly asking "Why?", structured with branching rules and prohibiting human error as a terminal cause. Forces engineering teams to look past surface symptoms and address the underlying organizational, architectural, or automation failures.