After-Action Review

Description

A short, structured debrief asking what was planned, what actually happened, why there was a difference, and what to sustain or improve.

The AAR was developed by the US Army to learn from training exercises. Its four questions (What was supposed to happen? What actually happened? Why was there a difference? What will we sustain or improve?) are answered by participants themselves, facilitated rather than judged, immediately after the event. Ground truth from instrumentation is used to correct memory. Its rigour comes from comparing intent with observed outcome, from rank-free participation, and from repetition across many events, which allows patterns to be aggregated. In UX it suits research operations (after each study), design sprints and launches, and also field studies with users who perform team activities.

Attributes

Loss type
Siloed
Effort
Hours

Steps

  1. Schedule the AAR immediately after the event, within hours, with all participants present.
  2. Restate the objectives and plan (what was supposed to happen) using written records.
  3. Reconstruct what actually happened using data, recordings and logs, letting participants add their perspectives.
  4. Discuss causes of differences, focusing on the system and decisions rather than individuals.
  5. Agree 2-3 items to sustain and 2-3 to improve, with owners.
  6. Store AARs in a searchable log and review trends quarterly.

Sources

  • US Army, TC 25-20: A Leader's Guide to After-Action Reviews (1993)
  • Darling, Parry & Moore, Learning in the thick of it (2005), Harvard Business Review

Also known as

  • AAR
  • Hot debrief

Categories

  • Research Operations
  • Research Design & Rigor