Simulations
Run a simulated survey against an AI-generated audience before you ever talk to a real customer.
Simulations build a synthetic panel of personas matching a target audience and run your questions against them, so you get directional answers in minutes instead of weeks.
The five-step wizard
Every Simulation walks through the same flow: Setup (name and type) → Target Audience (synthetic, real customers, or both) → Questions (manual or AI-generated) → Model Settings (persona count and response style) → Generation.
Frequently asked questions
What does a Simulation give me that real research can't as quickly?
A Simulation builds a synthetic panel of personas matching your target audience and runs your questions against them, giving you directional answers in minutes instead of weeks.
What are the five steps of the Simulation wizard?
Setup (name and type), Target Audience (synthetic, real customers, or both), Questions (manual or AI-generated), Model Settings (persona count and response style), and Generation.
Can a Simulation be based on my real customers instead of synthetic personas?
Yes — the Target Audience step lets you build the panel from synthetic personas, real customers already in your workspace, or a mix of both.
Articles
- Simulations overviewRun a simulation against an AI audience and get directional answers in minutes.
- Creating a SimulationWalk through the five-step wizard — setup, audience, questions, model settings, and generation.
- Building your target audienceDefine who your Simulation's simulated respondents should represent, or base them on your real customers.
- Reviewing simulation resultsRead your Simulation's overview, demographic breakdown, statistics, and individual responses.
- Managing saved audiencesReuse a target audience across multiple Simulations instead of rebuilding it each time.