After-Action Review
also known as AAR, structured debrief, post-action review, lessons-learned session
A structured facilitated discussion held immediately after a significant event, in which participants — not the facilitator — discover for themselves what happened, why it happened, and what to sustain or change. Developed by the US Army following analysis of the 1973 Yom Kippur War and institutionalised at the National Training Center from 1981. The facilitator guides through questions; rank is suspended; conclusions are owned by participants, not imposed by leaders. AARs are not critiques — they do not assign blame or determine winners and losers.
How the learner advances
Intent. Extract transferable lessons from a completed event by guiding all participants to discover — through structured questioning — what happened, why it happened, and how performance should change.
When to apply. Apply immediately after any significant event, exercise, project milestone, or failure where performance is worth understanding — while details are fresh and participants can still recall their decisions and reasoning. Also apply after successes: the discipline of asking why something went well is as valuable as asking why it went wrong, and less common.
Threshold — earns the next step. The team can name at least one specific practice they will sustain and one they will change, with a reason derived from the event — not a general resolve to 'do better' but a named behaviour tied to a named moment in the event.
Masterpiece — the artifact that proves it. A written lessons-learned record from the AAR that is specific enough to brief a team preparing for the next similar event — covering what was planned, what happened, why, and what concrete practices will be sustained or changed.
Facets
- Container — async
- Mode — reflectivecollaborativestructured
- Reach — team
- Persona — learnerpractitionerteam-member
- Craft (AI Fluency) — discernmentdiligence
- Learner — humanautonomous-agent
- Trainer — human
Inputs
- Completed event with participants — A training exercise, operation, project milestone, product launch, incident, or any significant shared activity that has recently concluded, with the people who participated available to reflect.
- Facilitator with the four AAR questions — A facilitator (not necessarily a leader) who guides the discussion through the four questions without imposing conclusions. The facilitator's role is to draw out participant discovery, not to deliver a verdict.
Outputs
- More capable team — A team that understands what drove its performance in the event — both strengths to sustain and gaps to close — and has shared ownership of the lessons because participants discovered them together.
- Lessons-learned record — A written record — the masterpiece — of what was planned, what happened, what caused the gap, and what the team will sustain or improve — actionable enough to inform the next similar event.
Steps (4)
Review what was planned
Ask: 'What were we trying to do? What was the intent?' Establish a shared baseline of the original plan, goals, and expected outcomes — without which the gap between plan and reality cannot be seen clearly.
Establish what actually happened
Ask: 'What actually happened?' Reconstruct the sequence of events as participants experienced it, not as leaders hoped it went. Use multiple perspectives — different participants saw different things. Avoid judgment at this stage; the goal is an accurate picture.
Determine why the gap exists
Ask: 'Why was there a difference between what was planned and what happened?' Facilitate discovery of causes — decisions made, information missing, coordination failures, unexpected conditions. The facilitator asks questions; participants surface the reasons. No blame is assigned.
Identify what to sustain and what to improve
Ask: 'What will we sustain and what will we improve?' Both questions matter equally. Participants name specific practices to keep and specific changes to make. The facilitator helps the team convert vague lessons ('communicate better') into concrete actions ('designate a decision log owner for the next event').
Principles
- Participants discover; the facilitator guides — conclusions imposed by a leader are not lessons learned, they are directives.
- Rank is suspended — the soldier who saw the failure must be as free to speak as the general who ordered the action; otherwise the AAR produces the leader's version, not the truth.
- Both successes and failures deserve an AAR — what went right is as worth understanding as what went wrong.
Known uses (3)
US Army National Training Center — US Army
military training FM 25-101: Battle Focused Training, Appendix G. Official Army doctrine definition. AAR institutionalised at National Training Center, Fort Irwin, from 1981.
Software engineering retrospectives — Agile development community
software engineering Sprint retrospectives in Scrum are a direct adaptation of AAR structure: what went well, what did not, what to change — same four questions in a software delivery context.
autonomous-agent training Post-episode structured review in agent training environments — what was the policy's intent, what trajectory resulted, what caused deviation, what parameter or prompt to adjust — mirrors the AAR four-question structure.
Known failure modes (2)
- [critique-masquerading-as-AAR]
Anti-pattern: the leader assigns blame, rates performance, or announces conclusions rather than guiding participant discovery. A critique delivered in AAR format is not an AAR — it produces compliance, not learning, and shuts down the honest reporting that makes AARs valuable.
- [lessons-without-specificity]
Anti-pattern: the AAR produces general resolve — 'we need to communicate better', 'coordination was an issue' — without naming specific behaviours to sustain or change. Vague lessons are not actionable and will not change the next performance.
Related trainings (4)
- Reflective Practice★★
Surface and revise the tacit knowledge driving professional performance by reflecting both during and after action.
- Experiential Learning Cycle★★
Deepen learning by cycling continuously through doing, reflecting, concluding, and experimenting rather than treating any single stage as sufficient.
- Experimental Exploration with Checkpoints★★
Resolve a specific uncertainty through a strictly time-boxed exploration so that the next planning or learning decision can be made on evidence rather than assumption.
- Learning by Doing★★
Produce genuine learning by immersing the learner in purposeful activity on a real problem where thinking is required and success is visible.
Sources (3)
“An AAR is a review of training that allows soldiers, leaders, and units to discover for themselves what happened during the training and why.”
“a method of providing feedback to units by involving participants in the training diagnostic process in order to increase and reinforce learning. The AAR leader guides participants in identifying deficiencies and seeking solutions.”
Provenance
- Ecosystem: military training, professional development, software engineering, autonomous-agent training
- Added to catalog:
- Last updated:
- Verification status: verified