AAR (After Action Review)
The After Action Review is four blunt questions asked right after something happens - what was supposed to happen, what actually happened, why the difference, what to do differently - so a team learns from events while the memory is still fresh instead of writing a report nobody reads later.
Blunt questions run in a row, each built directly on the answer before it.
Reach for this when…
- A project just ended and the team is about to scatter without ever debriefing it.
- The same mistake keeps recurring across projects because nobody writes down why.
- You want a fast, no-blame way to learn from something that just went wrong, or right.
How to run it
- Gather the team immediately after the event or project, while memory is fresh.
- Ask: what was supposed to happen?
- Ask: what actually happened, facts only?
- Ask: why was there a difference, no blame?
- Ask: what do we do differently next time, specifically?
- Write it down and feed it into the next planning cycle.
A worked example
Situation. Aissatou Diop ran Teranga Health Access, an NGO delivering mobile clinics across rural communities near Dakar, Senegal, and her teams moved straight from one outreach to the next without ever looking back.
Applied. After a vaccination drive fell short of its target, she called the field team together the same evening and ran the four questions before anyone went home.
Result. The gap turned out to be a supply-truck schedule nobody had checked against local market days, not staff performance as everyone had assumed going in. The next drive was planned around market days and hit its target.
The catch
It only works if 'no blame' is real, not a slogan - the first time someone gets quietly punished for an honest answer, future AARs get careful and useless. It captures what happened close to the event, but a stack of unread AAR notes helps nobody; someone has to actually carry the learning into the next plan. Run too soon after a genuinely painful event, it can feel like an interrogation rather than a debrief.
If the same finding shows up in your third AAR running, the problem isn't the event. It's that nobody's using the first two.
Origin: US Army; adapted by CCL and others