After Action Review: The Army's System for Learning from Every Experience

The US Army developed the After Action Review (AAR) following analysis of the 1973 Yom Kippur War. Institutionalized through the National Training Center at Fort Irwin (first training rotations 1981–1982), the AAR became the Army's primary learning mechanism: four questions asked after every significant event. Harvard Business School research shows it consistently outperforms post-training reflection alone.

2026-02-19 6 min read

Origins: From Yom Kippur to Fort Irwin

The US Army developed the After Action Review in response to performance analysis conducted after the 1973 Yom Kippur War between Israel and the Arab coalition. Analysis of US Army training and readiness identified significant gaps between how units planned to perform and how they actually performed under combat conditions. The Army began developing systematic methods for closing this gap.

The National Training Center at Fort Irwin, California was activated in October 1980. Its first full training rotations began in 1981–1982. The NTC provided a realistic training environment where units rotated through exercises against a permanent opposing force (OPFOR), with the specific goal of surfacing performance gaps that could be addressed through the AAR process. The NTC made the AAR a central feature of Army training doctrine, which was later formalized in training manuals including FM 7-0 and FM 7-1.

Four questions

The AAR structure is four questions applied after every significant event: (1) What was supposed to happen? (2) What actually happened? (3) Why was there a difference? (4) What should we sustain or improve? The simplicity is intentional, because complex formats would reduce usage frequency.

US Army After Action Review, FM 7-0 and FM 7-1. National Training Center at Fort Irwin, California.

The Four-Question Structure

The AAR structure is deliberately simple: four questions applied after every significant event, by the participants themselves rather than by outside observers.

What was supposed to happen? This question surfaces the intent: the plan, the expected outcome, the standard being measured against. It is necessary because teams often have different understandings of what success looked like, and the gap between actual and intended cannot be examined without a shared reference.
What actually happened? A factual account of what occurred: not evaluative, not attributive, but descriptive. The Army emphasizes that this phase should be treated as evidence collection: what were the observable events, in what sequence, with what consequences? Premature evaluation at this stage contaminates the factual account.
Why was there a difference? Causal analysis: what factors explain the gap between intended and actual? This is where root causes are identified, not just symptoms. The Five Whys technique is often used here to prevent stopping at surface-level explanations.
What should we sustain or improve? Action items in two categories: what worked that should be explicitly repeated, and what didn’t work that requires change. Both are important, since sustain items are as valuable as improvement items, because they document what to replicate.

A critical feature of the Army AAR is rank irrelevance: the review is conducted by all participants, and any participant can identify what went wrong regardless of who made the relevant decision. The review does not protect decisions based on who made them. This feature requires psychological safety to function. It works in the Army because it is structurally required, not optional.

Professional Applications

Project post-mortems and retrospectives. The AAR format is the model for agile retrospectives and project post-mortems. The four-question structure focuses the review on specific, actionable gaps rather than general impressions of how the project went. The most common failure mode of organizational post-mortems, attributing outcomes to individual performance rather than systemic factors, is countered by the question structure, which drives toward causal factors rather than blame attribution.
Individual performance reviews. The AAR structure applies to individual performance conversations: What was the plan? What happened? Why the difference? What changes? This framing removes the evaluative loading from performance discussions by treating them as evidence analysis rather than judgment. It also applies to self-reflection: a brief personal AAR after each significant presentation, meeting, or project completion builds a systematic learning practice from experience.
Meeting debriefs. Brief AARs after important meetings (particularly those with complex stakeholders or significant decisions) close the feedback loop that most organizations leave open. Two minutes of structured reflection (What was intended? What happened? Why the gap? What changes next time?) converts a completed meeting from a one-time event into a learning data point.

Frequently Asked Questions

How frequently should AARs be conducted?

The Army conducts AARs after every significant training event, which at the NTC can be multiple times per day. For organizational use, frequency should match the learning opportunity density of the work. High-stakes, high-frequency work (client-facing interactions, sales calls, product launches) benefits from more frequent AARs. The risk of excessive AAR frequency is fatigue and formulaic execution; the risk of too-infrequent AARs is that learning opportunities pass without capture. A practical calibration: conduct formal AARs after every significant milestone, important meeting, or project completion, and brief informal AARs (two to three minutes of structured reflection) after routine but important recurring events. The key variable is whether the AAR produces actionable changes. If the action items from AARs are consistently not implemented, frequency is less important than implementation quality.

What makes an AAR fail in practice?

Three common failure modes: (1) Blame attribution: the review degenerates into finding who made the wrong decision rather than why the outcome differed from the plan. This requires a specific facilitation norm: 'what factors caused this?' rather than 'who is responsible for this?' (2) Lack of follow-through: action items from the AAR are documented but not assigned, scheduled, or tracked, so they don't change behavior in subsequent events. AARs without accountability for action items are documentation exercises, not learning systems. (3) Missing the psychological safety requirement: in hierarchical organizations, lower-ranked participants may be unwilling to identify what went wrong if senior leaders made the relevant decisions. The Army's solution (structural requirement that rank is irrelevant) is difficult to replicate in civilian organizations; the best substitute is explicit facilitation norms and senior leader modeling of self-critique.

Is there research supporting the AAR's effectiveness beyond military contexts?

Yes. Bradley Staats, Francesca Gino, and Amy Edmondson conducted research on reflection and learning in organizational settings. Staats and Gino (2014, Academy of Management Journal) found that allocating time for reflection after task completion significantly improved learning rates in a call center context. Di Stefano, Gino, Pisano & Staats (2016) ran experiments showing that 15 minutes of structured reflection after a task produced 23% better performance on a subsequent task compared to a control group that used the same time for additional practice. The reflection advantage was mediated by improved self-efficacy: the reflection group gained confidence from explicating what they had learned, not just from additional experience.