The Incident Follow-Up Completion Rate

Postmortems produce action items. How many actually ship? The metric, the dashboard, and the conversation it forces.

The metric

The headline metric is per-quarter postmortem action-item completion rate. Healthy teams ship 70 percent or more on time; below 50 percent, the postmortem ritual continues but the system stops improving. Tracked per team because some teams ship and some do not.

• Per-quarter opened versus completed. Action-item ratio per team. Foundation metric.
• Healthy: 70 percent or higher. Standard bar per team. Below 50 means postmortems are theatre.
• Track per team. Granular metric per team. Some teams ship; some do not. Patterns reveal capability gaps.
• Published trend per team. Visible chart per team. Supports honest leadership conversations.

Dashboard

The dashboard is the visibility layer. Per-team counts (opened, completed, overdue, rate), leadership-visible, linked from every postmortem so authors see the team’s bandwidth before adding more items.

• Per-team four-count view. Opened, completed, overdue, rate. Colour-coded.
• Visible to leadership. Executive read-cadence. The metric stops being optional when leadership reads it weekly.
• Linked from every postmortem. Bandwidth check per postmortem. Authors see capacity before committing more items.
• Named dashboard owner. Responsible analyst per org. Catches stale or wrong dashboards.

The conversation

The metric drives resource decisions, not naming-and-shaming. Below 50 percent means a team needs help (action items too ambitious, or team genuinely overloaded); honesty about capacity beats committing more than the team can ship.

• Below 50 percent: team needs help. Underlying-cause investigation per team. Ambitious items or overloaded team.
• Resource decisions follow. Staff-up or stop-committing decision per team. Drives explicit choice.
• Honesty beats theatre. Fewer items shipped beats more items opened and dropped. Drives the right behaviour.
• Named action per quarter. Documented “what we will do” decision per team. Catches “we discussed but did not act.”