Design the Review Before the Session Ends

Learning objective

By the end of this lesson, you can define how a team will test both the use of the session artifact and the result it was intended to influence.

Why it matters

Positive energy, attendance and satisfaction can help diagnose the experience, but they do not show that the session changed the operating system.

A session is not complete when participants leave. It is complete when its output enters work and the organisation can inspect what happened next.

Core concept: review three levels

1. Artifact use: was the decision, charter or plan actually used?
2. Behaviour or mechanism: did the agreed handoff, control or decision rule change in practice?
3. Outcome: did the relevant customer, operational or risk measure move while guardrails remained acceptable?

Define the baseline, owner, evidence source and review date before participants leave. Use early process indicators for the first check and outcome indicators only at a horizon where change could reasonably appear.

Do not claim causality from one before-and-after number when other conditions changed.

Visual explanation

Text alternative: review moves through three levels—artifact use, mechanism change and outcome movement within guardrails—before the team retains, changes, stops or expands the intervention and feeds evidence into the next mandate.

Worked example: NordForge

• 30 days: pilot governance is active, site owners are named and the new handoff rule is used in sampled orders.
• 60 days: exception age, escalation volume and expedited freight are reviewed against baseline and guardrails.
• 90 days: OTIF movement, transferability and capacity evidence determine whether to continue, change, stop or expand the pilot.

The review does not ask only “Did we complete the actions?” It asks whether the mechanism was used and whether the evidence supports the next decision.

After Action Review as a learning loop

The U.S. Army's AAR sequence moves from what was expected and what happened to what should change next. Research syntheses of structured debriefs also report a positive average relationship with performance, while outcomes depend on the quality and context of the debrief.

Use six questions:

1. What did we expect?
2. What actually happened?
3. What explains the difference?
4. What should we retain?
5. What concrete rule should change?
6. What will we test in the next attempt?

A list of abstract “lessons learned” is not yet a learning loop.

Common mistake

Mistake: measuring only the final business metric.

When the outcome has not moved, the team needs to know whether the intervention was used, whether the proposed mechanism was wrong or whether the measurement horizon was too short.

Quick check

OTIF has not changed after 30 days, but the pilot was not yet operating at both sites. What can the team conclude?

A. The session failed.
B. The pilot works.
C. Outcome evidence is not yet interpretable without checking implementation and mechanism.
D. Expand immediately.

Answer: C. Review artifact use and mechanism before attributing the outcome.

Practical prompt

Use the 30–60–90 Review to define an owner, baseline, evidence source, date and response rule at each horizon. Schedule the After Action Review before the working session closes.

Summary

• Evaluate use, mechanism and outcome separately.
• Define the baseline and evidence source in advance.
• Match indicators to a realistic time horizon.
• Change a concrete rule, not just the language of a lesson.
• Feed review evidence into the next mandate.

Sources

• U.S. Army, FM 7-0: Training (opens in a new tab), After Action Review agenda.
• Scott Tannenbaum and Christopher Cerasoli, Do Team and Individual Debriefs Enhance Performance? (opens in a new tab).

Path completion

Apply the full chain to the Decision Room practice case: mandate → evidence → participation → options → decision → artifact → review.

The path is complete when another person can inspect your design and answer what the session is authorised to do, what evidence shapes it, why each participant is involved, who closes the decision, what artifact enters the workflow and how the organisation will learn from the result.