Run sheet: the 30-minute incident rehearsal

Use this alongside the eight invented incident cards on the home page. The scenario is fictional and contains no customer data.

Preparation
Choose three roles: interviewer, participant and observer. Give the participant and observer the full card packet. Give the interviewer cards A, D and G. Each person needs a blank copy of the worksheet below.

Copyable worksheet

Code:
Time:
Reported action:
Source card:
What the participant saw or did:
What remains unknown:
Next neutral question:

Make one entry per reported event. If a time is missing, write "not established". Keep a participant's interpretation separate from the action they describe.

Timing
0-4 minutes: read the role instructions and set up the notes.
4-10 minutes: read the available cards and sketch an initial sequence.
10-18 minutes: conduct the interview and reveal cards as their events are reached.
18-25 minutes: compare the notes with the full packet.
25-30 minutes: write a three-sentence incident account and discuss what remains unknown.

Observer's checks
1. Can another reader find a source card for every timed step?
2. Did the interviewer supply a missing detail in a question?
3. Is the request for a warning labelled as a suggestion?
4. Does the note distinguish the blank attachment from an unproven cause?
5. Is the 17-minute handoff interval described as elapsed time rather than uninterrupted effort?

Facilitator key
The timed sequence is H, C, B, A, D, F.
Card E has no established time or position in that sequence.
Card G is a proposed solution, not a timed incident or evidence about other people.

The first file was sent at 09:10 and the replacement at 09:27, an elapsed interval of 17 minutes. The cards do not establish the export's root cause, the selected date range, the number of affected people or a monetary cost.

A careful three-sentence account
The participant was asked to send current open-request totals for North. They reported selecting a region, downloading and sending a file, discovering a blank attachment after a teammate's question, and sending a manually assembled replacement. The selected date range and the cause of the blank attachment remain unknown.

Second round
Rotate the roles. Ask the new interviewer to recover the sequence without reading the key aloud. Compare which questions produced an event, which produced an interpretation and which introduced a solution before the sequence was clear.

Version 1 - 7 November 2026.