NOTQIN — OT Track

After-Action Reviews

3 min read

After-Action Reviews (AAR)

Definition

A structured, brief, peer-driven review held immediately after any meaningful event (a mission, deployment, sprint, customer call, pilot) to extract lessons that improve future performance. Originated in the US Army as a tool to convert combat experience into doctrine; now widely used across industry under names like “retrospective,” “post-mortem,” “debrief.”

The four questions

The Army’s canonical format — terse on purpose. Same four every time:

  1. What was supposed to happen?
  2. What actually happened?
  3. Why was there a difference? (or why wasn’t there one)
  4. What can we learn and do differently next time?

That’s it. Should fit in 30–60 minutes for most events.

Why it works

  • Immediacy — held while memory is fresh, not after the next event obscures detail
  • Peer-driven — no rank in the room; the lowest-ranking person speaks first to prevent senior-voice anchoring
  • Blameless framing — focus on the system + decisions, not individual fault
  • Forces externalization of tacit observations into explicit lessons (the Externalization mode)
  • Bias-resistant — same four questions every time defeats the “let’s redefine success after the fact” trap

Examples in NOTQIN context

EventAAR usage
End of every Wave (every 2 weeks)Wave-end retro using the retro skill — already in PM cadence
After every discovery call (Wave I onwards)10-min self-AAR, captured in Discovery Calls/<co>-<date>.md
After first pilot site visitFull team AAR — likely 90 min given stakes
After Alignment MeetingAhmed runs a 1-person AAR on what he heard
After every lost dealAAR captured in 04_Strategy/Lost Deals/<co>.md (folder per memory, currently empty)
After production incidentBlameless post-mortem (the SRE-flavored AAR)

Common mistakes

  • Skipping when the event went well — wins teach as much as losses
  • Letting it slide into blame (“who did what wrong”)
  • Doing it weeks later — most insight is gone
  • Not actually changing behavior after — AAR without follow-through is theater
  • Senior person talks first — anchors the conversation
  • No written capture — lessons evaporate
  • Treating it as a status report instead of a learning exercise

How NOTQIN runs AARs

  • End of each wave (every 2 weeks) — retro skill triggers structured review, updates to MEMORY.md, captures lessons in Decision Log or relevant files
  • End of each discovery call — short capture in the call note, tagged for retrieval later
  • End of each pilot phase — formal AAR with the customer team if possible (turns it into a Socialization moment too)

See also

Graph View

Backlinks