fac-facilitator-retro
Turns a fifteen-minute, six-question debrief by the trainer or the delivery team into a half-page retro note and a table of timing corrections per activity, process only. Use for "run fac-facilitator-
SKILL.mdpacks/trainers-pack/skills/fac-facilitator-retro/SKILL.md
GitHubFacilitator Retro
When To Use
You run the same session next month and cannot remember what you swore to change. Use this the same day, alone or with your co-trainers, while the room is still fresh. It answers: what did this session teach us about the design, and what does the next design need to know?
When Not To Use
This is the team's inward look at process. For the questions participants answer after an activity, run Activity Debrief Questions. For participants' view of the session, run Session Feedback Form. For what goes to the person who asked for the session, run Sponsor Debrief; the retro note is never sent to the sponsor.
Inputs
- The plan you ran from, with planned times per activity (a session plan, an agenda or a run sheet)
- Your answers to the six questions, typed or dictated, however short
- The prepared moves you had ready for difficult moments, if you wrote any If you have none of this, I start from your memory of the day in a few lines and mark the output as a first draft.
Approach
Tannenbaum and Cerasoli's 2013 meta-analysis of debriefs found that structured debriefs focused on learning, not blame, improve later performance, and that structure and facilitation appear to matter. So the retro is six fixed questions in a fixed order, and every answer ends as a correction to the design. The failure it prevents: the debrief that turns into a long talk about one person in the room, and the next run opens with the same block starved of time.
Workflow
- Ask three questions: which session and when, are you answering alone or as a team, and do you have the plan with its planned times?
- Hand over the six questions as a sheet you answer in fifteen minutes, in this order: (1) plan against room, where did the session leave the plan and was leaving it right; (2) activities, which earned their time, which did not, and why (wrong activity, wrong slot, wrong split, wrong day); (3) wobbles, which prepared moves were used, did they work, what had no plan; (4) time, where it really went and which block was starved; (5) room and tools, what was missing, broken or unnecessary; (6) the one change if you ran it again tomorrow.
- Take the answers as they come. A tired team's honest ten minutes beats a thorough review nobody wanted. Leaving the plan well is craft, so the first answer is never read as a fault.
- Where an answer names or points at a person ("the director derailed block three"), rewrite it as the process lesson: what the design does next time when a senior voice redirects the room. Names stay out.
- Build the corrections table from the plan and the answers: planned time, real time, group size and what to change, one row per activity. Real times come from you; a time nobody remembers stays a placeholder.
- Write the half-page note: what worked, what to change, the one change. Anything bigger than process (a conduct issue, something a participant disclosed) is listed apart with the human conversation it belongs to, and left out of the note.
- Say where the note lives. Kept in a Project with shared memory (beta, on select plans), it is there when the next design starts; otherwise paste it at the top of the next design chat.
Output Format
# Facilitator Retro Note
Session: [name] | Date: [date] | Answered by: [trainer alone / delivery team] | Time taken: [minutes]
## What worked
- [specific: the activity or prepared move, and what it produced]
## What to change
- [specific: the block, the tool or the room, and the change to the design or logistics]
## The one change
[The single thing to change if this ran again tomorrow.]
## Corrections by activity
| Activity | Planned time | Real time | Group size | What to change |
|---|---|---|---|---|
| [activity] | [minutes] | [minutes] | [number] | [correction] |
## Outside the retro
- [what surfaced that is bigger than process] goes to [the right person], not into this note
## Decision
[Lead trainer name] decides by [date, before the next design starts] which corrections go into the next run, and confirms the note stays with the delivery team.
Done When
- All six questions were answered, even in one line each, and each answer shows up in the note or the table
- Every activity that ran has a row, with real times from the team or an honest placeholder
- The one change is a single change, not a list
- No line names or describes a participant
- You know where the note will be when the next design starts
Quality Bar
- Specific beats complete: "the trio exercise needs a longer slot with this group size" is a correction, "pacing" is not
- Rooms are weather and design is the craft; a hard session is not read as anyone's failure
- If you are beating yourself up, the note keeps what is learnable as a correction, not a confession
- Written so any member of the delivery team could read it
- Process only: no assessment of any participant, and no rating of the trainer.
Next
Run fac-sponsor-debrief (Sponsor Debrief) to take what is useful to the person who asked for the session.
About the makers
This pack is made by Polar Bear, a consultancy built by ex-McKinsey founders with a dream to make AI work for People, not instead of them. We help our clients build people systems and AI-first ways of working, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).