Postmortem Action Items With Deadlines That Stick

Action items without deadlines never ship. The deadline discipline that drives delivery.

Set the deadline

Deadlines get set at the postmortem, not in the week after. Owner plus deadline per item, set in the room. Default 30 days; variations require a documented reason so the deadline cannot quietly stretch to 90. Owner explicitly acknowledges in the postmortem document so "I did not agree to that" cannot land later.

• Owner and deadline in the room. Named pair per action. Set at the postmortem; no "we will figure it out later."
• Default 30 days. Standard window. Long enough to ship a real fix; short enough to keep urgency alive.
• Variations need documented reasons. "Why this is not 30 days" note per action. Catches sandbagging before it becomes the norm.
• Explicit owner acknowledgement. Owner signs off in the postmortem document. Catches "I did not agree to that" pushback later.

Track

Tracking is weekly. Per-team list of in-flight items with deadlines and ownership visible; approaching deadlines highlight 7 days out so teams react before they slip; weekly status update per item catches stalled work before it goes overdue.

• Weekly per-team list. In-flight action-items view per team. Deadlines and ownership visible to everyone.
• Approaching-deadline highlight. "Due in 7 days" tag per action. Teams have time to react before the date.
• Always-on dashboard. Action-item view per team. Manager visibility without asking.
• Weekly status update per action. Progress note per item. Catches stalled work before it becomes overdue.

Escalate

Overdue items automatically escalate to leadership review. The conversation lands at the executive level so the team gets either help or consequences. Repeated overdues adjust the team's reliability budget: chronic non-delivery means less feature work and more reliability investment until the action-item velocity returns.

• Overdue surfaces in leadership review. Executive-visible flag per overdue action. Drives the conversation that prevents the next outage.
• Repeated overdues adjust reliability budget. Consequence rule per team. Chronic non-delivery shifts feature/reliability allocation.
• Named escalation owner per team. Manager who handles overdue conversations. Catches "everyone-and-no-one" follow-up.
• Quarterly action-item retro. Delivery-rate review per quarter. Continuous improvement of the discipline itself.