Writing an After-Action Report That Actually Closes the Loop

The after-action report is where preparedness programs either compound or stall. Done well, it converts one exercise or incident into a set of specific changes that make the next one go better. Done the way most are done, it produces a PDF that everyone praises and nobody opens.

The difference is rarely writing quality. It is whether the report ends in owned, dated corrective actions and whether anyone tracks them.

Why most AARs change nothing

Three failure patterns account for nearly all of it.

The report is diplomatic. Written to avoid embarrassing anyone, it describes problems in language so soft that the problem disappears. "Interagency communication presented some challenges" tells a reader nothing and obligates no one.

The report has no owners. Recommendations are addressed to the organization rather than to a position. Work assigned to everyone is done by no one.

Nobody looks at it again. There is no corrective action tracking, no standing agenda item, and no review before the next exercise. Findings roll forward untouched until they show up again, identically worded, two years later.

Collecting data while it is still accurate

Memory degrades fast and it degrades in a particular direction: people remember the outcome and reconstruct the decisions as more orderly than they were. Collect within days, not weeks.

• The hot wash, held immediately, while participants are still together. Three questions: what worked, what did not, and what single change would help most.
• Participant feedback forms, collected before anyone leaves the room.
• Evaluator notes against the Exercise Evaluation Guides, tied to specific objectives rather than general impressions.
• Objective artifacts: activation and notification timestamps, situation reports, resource request logs, press releases, and EOC message traffic. These are the record that corrects everyone's recollection.
• Follow-up conversations with two or three people who were quiet in the room. They usually saw the most.

For a real incident, the same discipline applies with higher stakes, because the documentation you assemble also supports reimbursement and any subsequent review.

What separates a finding from an observation

An observation notes what happened. A finding states what was wrong, why it matters, and what should change. Only findings are actionable.

Findings should name positions, not people. The goal is a system that works regardless of who is in the chair.

Root cause, not symptom

Ask why until the answer is something you can fix. A late situation report is a symptom. Asking why produces a chain: no one owned it, because the annex assigns it to a function rather than a position, because the annex was written before the EOC was reorganized in 2023.

The corrective action attached to the symptom is "remind staff to submit reports promptly," which will not survive the next activation. The corrective action attached to the root cause is "revise Annex B to assign situation report compilation to the Planning Section Chief position, and update the EOC position checklist," which will.

In practice, most findings trace back to one of five causes: planning (the plan is silent, wrong, or contradictory), organization (no one owns it), equipment (the tool was missing or did not work), training (people did not know), or exercise (this was never practiced). Categorizing findings this way makes patterns visible across multiple exercises.

A structure that works

An AAR does not need to be long. A ten to twenty page report is plenty for most county tabletops. A workable structure:

1. Executive summary. One page. What was exercised, against which objectives, the two or three most important findings, and the overall assessment. Many readers will read only this, so make it carry weight.
2. Exercise overview. Dates, scope, objectives, participating organizations, scenario summary.
3. Analysis by objective. For each objective, what was demonstrated, strengths worth preserving, and areas for improvement with supporting detail.
4. Findings and recommendations. Each finding numbered, categorized by cause, with a recommended corrective action.
5. Improvement Plan. The table below.
6. Appendices. Participant list, feedback summary, acronyms.

Name strengths honestly and specifically. A report that is only criticism gets defended against rather than acted on, and genuine strengths are worth documenting so they survive staff turnover.

The improvement plan is the deliverable

Everything before this section is context. The improvement plan is what produces change, and it belongs in a table with five columns and no empty cells.

Three rules for this table. One owner per action, by position. A real date, negotiated with the owner rather than assigned to them. An action small enough to finish; "improve communications" is a program, not a corrective action, so break it into pieces someone can complete.

Tracking to closure

A corrective action program does not require software. A shared spreadsheet reviewed on a schedule outperforms an expensive system nobody opens. What it does require is a rhythm:

• A standing quarterly review of open items with the owners present.
• Evidence of closure. An item closes when the revised annex is adopted or the training roster shows the names, not when someone says it is done.
• Carry-forward visibility. Items that miss their date stay visible, with a new date and a reason. Silent slippage is how findings become permanent.
• A pre-exercise check. Before the next exercise, review open items and build at least one objective around whether a previously corrected gap has stayed corrected. This is the step that converts a series of exercises into a program.

State reviewers look for exactly this loop. Open findings that roll forward without dates or owners are among the most common observations in a program review, and they are the easiest finding to have already closed before anyone asks.

Common questions

How soon after an exercise should the AAR be delivered?

HSEEP practice points toward delivering the draft within roughly 30 days and finalizing after the improvement planning conference. The practical reason for speed is accuracy: the longer the gap, the more the record is reconstructed rather than remembered.

Do real incidents need an AAR too?

Yes, and they are more valuable than exercise AARs because the conditions were real. Many organizations skip them because everyone is exhausted and moving on to recovery. Capturing even a two-hour hot wash and a short improvement plan is far better than nothing.

Who should write the AAR?

Ideally someone who observed but did not participate, and who did not write the plan being evaluated. Participants write reports that defend their own decisions without intending to. If the plan author must write it, have an outside reviewer read the draft against the evaluator notes.

What is the difference between an AAR and an Improvement Plan?

The AAR is the analysis: what happened, measured against objectives. The Improvement Plan is the commitment: specific corrective actions with owners, dates, and status. They are usually published together as an AAR/IP, and the second half is the half that matters.

Should the AAR be public?

Practices vary by jurisdiction and state law, and portions may be sensitive. What matters more is that the improvement plan is visible to the people accountable for it. If concerns about publication are driving soft language, resolve the disclosure question first rather than writing a weaker report.