Gather materials for a Virtual Coach role-play

Use this reference to prepare the documents, context, and prompt details a Virtual Coach role-play needs before you start building.

The Virtual Coach AI assistant does the heavy lifting: it writes your scenario, builds your persona, and generates your evaluation criteria. To do that well, it needs to understand what you’re training for. The more context you bring, the better your role-play will be. Virtual Coach does not automatically use your organization’s existing Adobe Learning Manager course content. You must supply reference materials or a prompt for each role-play you create.

Documents to bring

The AI assistant can use any materials you share. The following table shows what each document type is most useful for.

Document type
Accepted formats
What it helps build
Call recordings or transcripts
MP4, WAV, WEBM
Scenario and conversation context
Playbooks, talk tracks, or objection-handling guides
PDF, DOCX
The persona’s concerns and objections
Product one-pagers or pitch decks
PPTX, PDF, DOCX
What the learner should pitch and explain
Buyer persona or ideal customer profiles
PDF, DOCX
The persona’s role, tone, and priorities
Scoring rubrics or coaching frameworks
PDF, DOCX, CSV
The evaluation criteria and grading

Each document type feeds a different part of the AI-generated role-play. So, bringing more than one type produces a more complete scenario.

NOTE
Accepted file types are PDF, DOCX, PPTX, TXT, CSV, MP4, WAV, and WEBM. You can upload up to 20 files per session, up to 200 MB per file. If you upload more than 20 files, only the first 20 are processed; the rest are ignored without a warning. Excel (.xlsx) files aren’t supported — export .xlsx content to CSV before uploading.
TIP
Even rough notes or bullet points help. You don’t need polished documents, just enough context for the AI to understand your scenario and training goals.

Four questions to think through

You don’t need full written answers to these questions. Just come with your thinking. The AI assistant asks follow-up questions and helps you fill in the gaps.

What’s the scenario?
What’s the business situation, and why are these two people talking? Think about the context, the relationship stage, and the purpose of the call. For example, a first discovery call, a follow-up after a demo, a renewal conversation, or an internal coaching session.

Who is the persona?
Who will the learner be speaking with? Think about the persona’s title, attitude, communication style, and what pressures or priorities shape how they show up. For example, a skeptical VP of Sales, a budget-constrained IT Director, or a collaborative but distracted HR leader.

What’s the goal of the conversation?
What does a successful conversation look like, and what should the learner accomplish by the end? Think about the outcome you’re training toward, not just the topics. For example, uncover pain, earn a follow-up meeting, present a tailored pitch, or handle a specific objection and advance the deal.

How will the learner be graded?
What are the three to six key skills or behaviors that define a great performance in this conversation? Think about specific, verbal actions you’d want to hear in a transcript, not attitude or personality traits. “Discussed value” is vague. “Explained how the platform reduces onboarding time by 40%” is gradable.

Check our design guide to learn more about designing a role-play.

Quick checklist before you start

  • You have at least one document, or a scenario you can describe, to share with the AI assistant.
  • You know the general scenario and relationship context.
  • You can describe the persona’s role, tone, and main concern.
  • You know what a successful outcome looks like for the learner.
  • You have a rough sense of the skills or behaviors you want to grade.

Best practice prompt template

Use this template as a starting point when you write your initial prompt for the AI assistant:

I want to build a single [conversation type] role-play focused on [skill, product, pitch, process, or business conversation]. The scenario should place the learner, a [learner role/title] at [learner company or team], in a [conversation setting] with [persona name or persona type], a [persona role/title] at [persona company or organization type]. The conversation is happening because [realistic business trigger, prior event, customer need, market context, or relationship stage].

The role-play should test whether the learner can [primary task they must perform naturally in the moment], while also [secondary skill or conversation behavior]. Strong performance should sound [describe desired communication style: natural, consultative, executive-ready, concise, empathetic, commercially relevant, and so on] and should include [key messages, discovery areas, proof points, questions, objection handling, or next-step behavior]. The simulation should reward [specific observable behaviors to encourage] and avoid rewarding [generic pitching, vague discovery, product dumping, over-talking, scripted language, skipping discovery, weak closing, and so on].

Example, using the product launch readiness use case: “I want to build a single multi-persona pitch role-play focused on positioning our new security add-on to an enterprise buying committee. The scenario should place the learner, an Account Executive at our company, in a scheduled solution-review meeting with four stakeholders at a mid-sized logistics company: a CFO focused on ROI, a Procurement Manager focused on contract terms, an IT Director focused on security and integration, and an end-user Champion focused on day-to-day usability. The conversation is happening because the customer is evaluating the add-on ahead of a contract renewal.”

Once you have your documents and answers ready, continue to create and publish a Virtual Coach role-play.

recommendation-more-help
learning-manager-help-migrated