Choosing the right roleplay creation method
Outdoo supports several ways to build a roleplay agent. They all end in the same place, a persona your team can practice against, but they start from different inputs and suit different situations. This article compares them so you can pick the right starting point instead of defaulting to the first option you see.
The Create Roleplay Agent menu is the entry point for every method below.
Comparing the methods
From existing calls
Upload a real call recording, or connect a source like Gong, and Outdoo extracts buyer personality, objections, tone, and communication patterns from it. Best for training that needs to feel exactly like your real customers. See Create a Roleplay Agent from Existing Calls.
From a call transcript
Similar to the calls method, but starting from a text transcript instead of an audio or video recording. Best for teams that already have transcripts on hand from a call intelligence tool but not the original recordings. See Create an AI Roleplay Agent from a Call Transcript.
From a prompt
Write a short description of the buyer, company, mindset, and call context, and Outdoo generates persona traits, objections, and conversation style from it. Best for fast, on-demand persona creation when you do not have a source call or document to build from. See Create a Roleplay Agent from a Prompt.
From a template
Start from a ready-made template built for common scenarios like cold calls, discovery, or demos. Best for structured, repeatable simulations and for new managers who want a proven starting point rather than building from scratch. See Create a Roleplay Agent Using Templates.
From a file or resource
Attach product documentation, a pricing guide, or a sales playbook from your Resources library, and the agent grounds its behavior and facts in that content. Best for factual accuracy and keeping the agent aligned to your specific messaging. See Create a Roleplay Agent from a File or Resource.
From a LinkedIn profile
Paste a public LinkedIn profile and Outdoo builds a persona around that person's job title, responsibilities, company background, seniority, and likely priorities. Best for outbound prospecting, account-based selling, and prepping for a specific named stakeholder. See Create a Roleplay Agent from a LinkedIn Profile.
Multi-persona, from scratch
Build an agent that plays multiple stakeholders in the same conversation rather than a single buyer. Best for simulating buying committees and multi-threaded deals where reps need to manage several viewpoints at once. See Create a Multi Persona Roleplay Agent.
In-person
Configure an agent for face-to-face practice rather than a phone or video call. Best for field sales, retail, or any role where the real conversation happens in person rather than over a call. See Create an In-Person Roleplay Agent.
Video avatar (video mode)
Run the roleplay against a visual, on-screen avatar instead of a voice-only or chat-only agent. Best for reps who need to practice reading and reacting to visual cues, or for video-first sales motions. See Create a Video Avatar (Video Mode) Roleplay Agent.
Micro-learning
Build an agent that focuses on one narrow slice of a conversation, such as objection handling or pricing negotiation, instead of a full end-to-end call. Best for short, targeted practice reps can complete in the flow of their day. See Create a Micro-Learning Roleplay Agent.
If you're not sure, start here
If you have a real call recording or transcript available, start from that, it produces the most realistic persona with the least manual setup. If you do not, a prompt is the fastest way to get something usable in minutes, and you can always add resources or clone the agent into other sales stages later.