liveBook · Manning

Cause analysis

As explained in Think Like a Software Engineering Manager

Root-cause analysis is a systematic method for identifying the underlying factors that produced a problem or failure.

A missed deadline or a failed deployment tells you something went wrong, but not why. Cause analysis investigates the factors behind the problem before you decide how to respond. For an engineering manager, that may mean examining an employee’s underperformance or learning from a production incident. section 4.5.1 section 12.2.12

Start with what you can observe, then investigate without jumping to blame:

  • For underperformance, notice changes in work quality, missed deadlines, or reduced engagement. Consider skill gaps, personal circumstances, role fit, resources, unclear expectations, and changing interests rather than assuming one explanation. section 4.5.1
  • For an incident, inform affected people, identify the underlying problem, assess customer impact, implement a fix, and document the troubleshooting and resolution. Use what you learn to choose actions that reduce recurrence. section 12.2.12

Ask why again

Suppose a system crashes. An initial answer is that its database server became unresponsive. Asking why again reveals that the server ran out of disk space; further questions trace the space use to large log files from a backup process that failed to rotate or clean them up. The 5 Whys approach repeats this question—often about five times, or deeper if needed—to get beneath the first explanation. section 12.2.12

Keep the inquiry focused on correction and learning, not a blame game. In a performance conversation, gather information objectively and give actionable feedback; after an incident, record what happened so the team can learn from it. section 4.5.1 section 12.2.12 section 1.1.3

How Think Like a Software Engineering Manager covers it

How deep it goes
Provides performance indicators and possible causes, plus an iterative 5 Whys incident example; stops short of a formal investigation protocol.
Examples
Uses a disengaged engineer missing deadlines and a system crash traced through database disk exhaustion to accumulating backup logs.
What it assumes
Assumes familiarity with workplace performance or production problems, while explaining the 5 Whys through a concrete incident.
What you can do after
Practice/perform it by checking plausible causes without bias, asking successive why questions, and turning findings into feedback or preventive actions.
How it connects
Connects one-on-one feedback and performance management with incident response, postmortems, and blameless learning.

Also explained in

Compare how 3 books explain cause analysis →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Think Like a Software Engineering Manager. The section links point to the parts of the book it draws on.