Training Session (Remote)
A 90-minute back-of-the-room session delivered over Teams with a Mural board as the room. The facilitator talks for less than fifteen minutes total. Everyone leaves having written a real play.
Designed on Sharon Bowman's 4Cs: Connections, Concepts, Concrete Practice, Conclusions. Remote back-of-the-room has one extra rule the in-person version does not need: instructions live on the board, not in the facilitator's voice. In breakout rooms nobody can hear you, so every activity zone on the Mural carries its own instructions, timings, and an example. If a breakout goes quiet and productive, the board did its job.
Logistics
- Time: 90 minutes including a 5-minute break at the midpoint
- Group: 8 to 24 people, breakout rooms of 4
- Platform: Teams for voice and breakouts, Mural for everything else. The deck is not screen-shared (see board spec)
- Facilitation: solo below 12 participants. Above 12, a producer handles breakouts, timer, and tech triage so the facilitator can visit rooms
- The one prep task that matters: collect two or three of the organisation's own documents that call themselves playbooks, guides, or ways of working, and drop them (or links) onto the board. The session runs on the org's own material, not hypotheticals. Pick documents whose authors are not attending, or brief the authors beforehand
Tech prep, one week out
- Send the Mural visitor link with the invite. Test it from a personal device outside the org network, because the session dies in minute one if the board is behind a login half the room cannot pass
- Pre-assign breakout rooms in Teams if the platform version allows it, and keep the same groups all session, re-forming rooms costs two minutes each time
- Ask for cameras on in the invite, and mean it for breakouts more than plenary. Do not police it live
- Have a fallback for the always-one person for whom Mural will not load: they pair with a room-mate and dictate
Board spec
Build the Mural as six zones left to right, so the session physically travels across the board and everyone can see where they are:
- Arrivals wall. The warm-up prompt in large text, blank space below for stickies
- Concepts strip. Slides 2, 3, 4, 5, 6, and 8 of the intro deck exported as images and placed in a row. No screen share: the facilitator talks over the board while people look at the same images they will annotate later. This also means late joiners see the whole arc at a glance
- Teach-back frames. Five frames, one per lifecycle stage, each preloaded with the text of that stage's page from the collection and a blank sticky labelled "the one thing" and another labelled "the antipattern, in our words"
- Test worksheets. One frame per breakout room, each containing the five questions of the Is It a Playbook test as a checklist, a link or embed of that room's assigned internal document, and a verdict box listing the neighbour genres
- Play frames. One frame per participant, name on it, preloaded with the Play Template sections as labelled areas. Private mode off, the point is that plays are visible
- Gallery and commitments. Empty at start. Voting session runs here, and a commitments row sits below it
Use Mural's outline feature so each zone is a numbered outline entry, and the facilitator can say "outline item 4" instead of talking people across the board. Lock every element that is not meant to move. Unlocked instructions get dragged by accident within minutes.
Connections - 12 minutes
Warm start (starts before the start). The Teams lobby opens 5 minutes early. The board link is in chat, and the arrivals wall prompt reads: "Think of a guide, manual, or process document you were given and quietly ignored. Add a sticky: one word for why." People who arrive early are already working.
Pairs in breakouts (5 min). Rooms of 4 open, but the instruction on the board says pair up within the room. Each person shares their ignored document and the why behind their sticky.
Harvest (4 min). Back in plenary. The facilitator drags stickies into clusters live on the board, thinking out loud, no lecture. Typical clusters: didn't fit my situation, out of date, too long, didn't trust it, couldn't find the bit I needed. One sentence to close: "Everything on this wall is a design failure we can prevent, and that is the next 75 minutes."
Remote note: dragging the stickies yourself while narrating is faster and holds attention better than asking the room to self-organise them. This is the one moment the facilitator performs.
Concepts - 20 minutes
Board walk (10 min). Zoom the board to the concepts strip and walk the six slide images in order: the two deaths, the anatomy of a play, the test and its neighbours, the core thesis, the lifecycle, the improvement loop. Connect each to a cluster from the arrivals wall. No screen share, everyone stays in the board they will spend the session in.
Teach-back jigsaw (10 min). Each breakout room takes one lifecycle frame. Six minutes to read the stage text and fill the two stickies: the one thing the rest must know, and the antipattern in their own words. Then plenary, 45 seconds per room, spokesperson reads their stickies with the frame zoomed. The facilitator adds nothing unless a room misses its stage's when-not-to.
Remote note: with fewer than five rooms, double up stages rather than skipping govern. Govern is never the stage to skip.
Break - 5 minutes
Timer on the board. Cameras off, stand up, leave the desk. Say the return time twice.
Concrete Practice - 38 minutes
Activity 1: run the test (15 min). Rooms open. Each room's test worksheet names its assigned internal document. The room reads it, answers the five questions in the checklist, and writes a verdict: either what genre it actually is and why that is fine, or the one change that would make it a playbook.
Debrief in plenary (4 of the 15): one verdict per room, spokesperson, worksheet zoomed. The facilitator's only job is protecting the honest verdict. A document that turns out to be a good runbook is a pass, not a fail. The point is the label, not the genre.
Activity 2: write a play (23 min). Solo first (10 min), everyone stays in plenary but muted, working in their own play frame. Each person picks an action from their own practice, something done at least three times that worked, and fills the template. Entry conditions first, then when-not-to, then the rest. The extraction rule is written at the top of every play frame: what happened, not what should happen, and keep the embarrassing detail in.
Pairs in breakouts (8 min): rooms of 4 open, pair up, read your partner's play frame. The reader's only question: "could I run this on Monday without asking you anything?" Every question the reader has to ask is a gap. Readers add small stickies to the play frame marking gaps. Writers do not defend, they collect.
Silent sweep (5 min): back in plenary, muted, everyone revises their play against the stickies.
Facilitator watch-fors, remote edition:
- Empty play frames at minute 4 of the solo block. Visit by direct chat, not by broadcast: "stuck on picking an action, or stuck on the template?"
- Plays about things done once. Ask in chat: "when was the third time?" and let them switch topics without ceremony
- Values in play costume: "communicate early and often" is a wish. Chat: "what would I see you do?"
- One person in a breakout doing all the annotating. The producer flags it, the facilitator drops into the room
Conclusions - 15 minutes
Gallery vote (7 min). All play frames are the gallery. Start a Mural voting session: everyone gets two votes for "the play I would steal." Silent reading and voting, no discussion. Remote galleries are faster than physical ones, so use the spare minutes as reading time, the plays deserve it.
Commitments (5 min). Each person adds a card to the commitments row: the next thing they will do with this, by when. The honest options: start a candidate list, finish this play and show it to the person it names, or run the test on a document they own. Cards stay visible, they are not collected or read out.
Close (3 min). The facilitator names the two or three most-voted plays by title only, then points at the canonical collection and ends with: "Today's plays are candidates. They graduate when someone who isn't you runs one and it works. Go find that person. The board stays open for a week."
After the session
Export the board to PDF before touching anything. The voted plays are the organisation's candidate list, started with demand evidence attached. If the session is part of establishing a collection, the most-voted plays are the first extraction targets, and their authors are the first people to receive the legitimise-the-writing permission for real.
Leave the board open for a week with edit access, and say so in a follow-up message. Some of the best play revisions arrive on day three, once people have had time to notice the missing when-not-to. Then archive it read-only, because a workshop board is a projection too, and the plays that matter move into the collection with candidate status and their author's name.
Run the session again in eight weeks with a different group. If none of the first session's plays have been run by a stranger by then, that is a signal about the org, not the session, and it is exactly the signal the govern stage exists to read.
This guide is provisional and follows the collection's rules: the first three facilitations are the test suite. Remote timings especially, breakout transitions in Teams cost 60 to 90 seconds each and there are five of them, which is why this version carries five fewer minutes of activity than the session appears to contain.
Licensed under CC BY-SA 4.0.