Action vs Blame

System framing.

Overview

Action vs blame is the discipline of framing postmortems around system improvements rather than individual fault. Blame produces fear; fear produces silent incidents that nobody documents and nobody fixes. Action framing inverts the trade: people share what actually happened because the document is about the system, not them, and the team can fix the system.

• System framing. What in the system allowed this to happen and propagate. Catches the design and process gaps that individual blame hides.
• Improvement focus. What concrete change prevents recurrence. Action items, owners, dates.
• Psychological safety. People share what they actually saw and did. Honest PMs are the only useful kind.
• Learning culture plus cross-PM patterns. Mistakes become inputs to improvement; action-framed PMs aggregate into systemic patterns nobody sees in any single incident.

The approach

System language, no individual naming, focus on changes, train new on-callers in the framing, document the policy. The discipline is mechanical, but the culture it produces is the asset.

• System language. "The system allowed X" rather than "the engineer caused X." Neutral framing by construction.
• No individual naming. Roles and actions, not names. Preserves safety even on the most painful incidents.
• Focus on system changes. What change prevents recurrence; who owns it; when does it land. Real follow-through, not platitudes.
• Train new on-callers plus documented template. Blameless framing has to be taught explicitly; the team's PM template enforces neutral language so consistency does not depend on memory.

Why this compounds

Each blameless PM compounds the next one. Engineers share what they saw; system patterns surface across incidents; real fixes reduce recurrence; the culture reinforces itself. By year two, blameless framing is the team's default mode of analysis, not a deliberate choice.

• Better PM quality. Action framing produces real root causes. Learning is the output.
• Better psychological safety. Engineers share what really happened. Honest PMs follow.
• Lower incident rate. Real fixes reduce recurrence. Uptime improves as a side effect.
• Year-one investment, year-two habit. First PMs establish the framing; subsequent ones run on the template.