Build the Role-Play
When To Use
People sit through role-play, go out to real customers, and you hope it stuck. Use this when the skill is a live conversation (a customer call, a hard message, a first coaching talk) and people need to practise it with each other. It settles what exactly they practise, what good sounds like, and what each person takes back to work.
When Not To Use
If the skill is a decision made alone, on screen or on paper, use Write the Scenarios. To prepare the facilitators who will run these rounds, use Train the Trainers. For the full manager kit, use Brief the Managers; here managers get one short note. If the group does not yet trust each other, start with the two models and practice in threes before anyone plays in front of the room.
Inputs
- The conversation to practise, the objective, and the learners' role.
- Real situations from the job: what the other party tends to want, where it usually goes wrong, and whether managers will get the short note.
If you have none of this, I start from the conversation and the role, and mark the output as a first draft.
Approach
The method is behaviour modelling training, from the meta-analysis by Taylor, Russ-Eft and Chan (2005): skill grew most with clear learning points, and transfer was greatest with mixed good and poor models, scenarios trainees brought themselves, goal setting and prepared managers. Effects shrink from the room to the job, so the kit plans the carry-over instead of hoping for it. The failure this prevents: two volunteers improvise at the front, the room says "that was good", and nobody can name what to do differently on Monday.
Workflow
- Ask at most three questions: which conversation, what goes wrong in it today, and whether managers will get the short note.
- Write four to six learning points, each short enough to hold mid-conversation. If a point needs a second sentence, it is two points or it is content.
- Write mixed models: one good and one poor version of the SAME conversation, each marked with the points it shows or misses. The poor one is believable, not a caricature.
- Write role cards for three roles (learner, other party, observer): the situation, what the other party wants, how hard to push. One prepared round, then rounds on scenarios the learners bring from their own work. Wording comes from your inputs; I invent no customer quotes.
- Write the observer note against the learning points only: saw, worked, one thing to try. It is handed to the learner and to nobody else. If you ask for a scoring sheet on individuals, I decline and offer this note.
- Write the debrief questions, then the goal slip: each learner writes what they will try, with whom and when, and keeps it. Draft the short note for managers: the learning points and one question to ask, with no report on any person.
Output Format
# Role-Play Kit
Conversation practised: [what] | Learners' role: [role] | Objective: [what people will do at work]
## Learning points
1. [short rule] 2. [short rule] 3. [short rule] 4. [short rule]
## Two models of the same conversation
| Model | Outline | Points shown | Points missed |
|---|---|---|---|
| Good | [outline] | [numbers] | [numbers] |
| Poor | [outline] | [numbers] | [numbers] |
## Role cards
| Round | Learner (situation) | Other party (wants, how hard to push) | Observer watches for |
|---|---|---|---|
| Prepared | [situation] | [wants, difficulty] | [learning point numbers] |
| Own scenario | [brought by the learner] | [briefed by the learner] | [learning point numbers] |
## Observer note (handed to the learner only)
Saw: [what happened] | Worked: [point] | One thing to try: [point]
## Debrief and goal
Questions: [what happened] / [what worked] / [what it means on the job]
My goal at work (kept by the learner): I will try [what], with [whom], by [when].
## Note for managers
Learning points: [list]. One question to ask: [question]. No report on any person follows.
## Decision
[Trainer name] agrees the learning points and both models with [expert name] by [date], and decides by [date] whether managers get the note.
Done When
- The learning points are short enough to recall mid-conversation.
- Both models show the same conversation and are marked against the points.
- At least one round runs on scenarios the learners bring, and every learner leaves with their own written goal, kept by them.
Quality Bar
- Role cards sound like the real job, drawn from the inputs.
- The observer note describes behaviour against the points, never a judgement of the person.
- The learner decides what to share with anyone, including their manager.
- Observer notes go only to the learner; no role-play is scored, recorded against a person or reported to a manager.
Next
Run ldg-check-the-ai-draft (Check the AI Draft) before any of this reaches a learner.
About the makers
This pack is made by 90n (90n.ai): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).
1---2name: ldg-build-the-role-play3description: Builds a role-play kit with learning points, role cards, a good and a poor model of the same conversation, observer notes that go only to the learner, debrief questions and each learner's own goal for the job. Use for "run ldg-build-the-role-play", "role play training", "role play scenarios", "role play exercise", "practise a difficult conversation", "sales role play", "observer notes for role play", "debrief questions after role play", part of The Complete Claude Guide for L&D Pack by 90n.4metadata:5 source: https://skillmd.com/skills/polar-bear-org/ldg-build-the-role-play6---78# Build the Role-Play910## When To Use11People sit through role-play, go out to real customers, and you hope it stuck. Use this when the skill is a live conversation (a customer call, a hard message, a first coaching talk) and people need to practise it with each other. It settles what exactly they practise, what good sounds like, and what each person takes back to work.1213## When Not To Use14If the skill is a decision made alone, on screen or on paper, use Write the Scenarios. To prepare the facilitators who will run these rounds, use Train the Trainers. For the full manager kit, use Brief the Managers; here managers get one short note. If the group does not yet trust each other, start with the two models and practice in threes before anyone plays in front of the room.1516## Inputs17- The conversation to practise, the objective, and the learners' role.18- Real situations from the job: what the other party tends to want, where it usually goes wrong, and whether managers will get the short note.19If you have none of this, I start from the conversation and the role, and mark the output as a first draft.2021## Approach22The method is behaviour modelling training, from the meta-analysis by Taylor, Russ-Eft and Chan (2005): skill grew most with clear learning points, and transfer was greatest with mixed good and poor models, scenarios trainees brought themselves, goal setting and prepared managers. Effects shrink from the room to the job, so the kit plans the carry-over instead of hoping for it. The failure this prevents: two volunteers improvise at the front, the room says "that was good", and nobody can name what to do differently on Monday.2324## Workflow251. Ask at most three questions: which conversation, what goes wrong in it today, and whether managers will get the short note.262. Write four to six learning points, each short enough to hold mid-conversation. If a point needs a second sentence, it is two points or it is content.273. Write mixed models: one good and one poor version of the SAME conversation, each marked with the points it shows or misses. The poor one is believable, not a caricature.284. Write role cards for three roles (learner, other party, observer): the situation, what the other party wants, how hard to push. One prepared round, then rounds on scenarios the learners bring from their own work. Wording comes from your inputs; I invent no customer quotes.295. Write the observer note against the learning points only: saw, worked, one thing to try. It is handed to the learner and to nobody else. If you ask for a scoring sheet on individuals, I decline and offer this note.306. Write the debrief questions, then the goal slip: each learner writes what they will try, with whom and when, and keeps it. Draft the short note for managers: the learning points and one question to ask, with no report on any person.3132## Output Format33```markdown34# Role-Play Kit35Conversation practised: [what] | Learners' role: [role] | Objective: [what people will do at work]36## Learning points371. [short rule] 2. [short rule] 3. [short rule] 4. [short rule]38## Two models of the same conversation39| Model | Outline | Points shown | Points missed |40|---|---|---|---|41| Good | [outline] | [numbers] | [numbers] |42| Poor | [outline] | [numbers] | [numbers] |43## Role cards44| Round | Learner (situation) | Other party (wants, how hard to push) | Observer watches for |45|---|---|---|---|46| Prepared | [situation] | [wants, difficulty] | [learning point numbers] |47| Own scenario | [brought by the learner] | [briefed by the learner] | [learning point numbers] |48## Observer note (handed to the learner only)49Saw: [what happened] | Worked: [point] | One thing to try: [point]50## Debrief and goal51Questions: [what happened] / [what worked] / [what it means on the job]52My goal at work (kept by the learner): I will try [what], with [whom], by [when].53## Note for managers54Learning points: [list]. One question to ask: [question]. No report on any person follows.55## Decision56[Trainer name] agrees the learning points and both models with [expert name] by [date], and decides by [date] whether managers get the note.57```5859## Done When60- The learning points are short enough to recall mid-conversation.61- Both models show the same conversation and are marked against the points.62- At least one round runs on scenarios the learners bring, and every learner leaves with their own written goal, kept by them.6364## Quality Bar65- Role cards sound like the real job, drawn from the inputs.66- The observer note describes behaviour against the points, never a judgement of the person.67- The learner decides what to share with anyone, including their manager.68- Observer notes go only to the learner; no role-play is scored, recorded against a person or reported to a manager.6970## Next71Run ldg-check-the-ai-draft (Check the AI Draft) before any of this reaches a learner.7273## About the makers7475This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).