Surveys

Surveys in AudiAInce generate model-based, survey-style responses across 19 structured AI audience archetypes. They can help you compare options, surface objections, and prepare questions for real-customer research. The archetypes are not human respondents, and the output is not a substitute for a representative survey.

Survey Types

Product Survey

Explore model-generated reactions to:

overall interest and appeal
feature priorities
price perceptions
stated purchase intent
open-ended questions

Concept Test

Compare ideas before building them:

concept clarity
perceived distinctiveness
value-proposition reactions
questions and improvement suggestions

Competitive Comparison

Explore how archetypes respond to products side by side:

stated preference
attribute-by-attribute reactions
switching objections
segment-level differences inside the model

These outputs do not measure market share or predict conversion.

Creating a Survey

Go to Survey in the app navigation.
Select your product.
Choose a survey type.
Configure questions or generate a draft with AI.
Run the survey-style simulation.

AI-Generated Questions

AudiAInce can draft questions from your product and research goal. Review wording, bias, response options, and suitability before using those questions with customers or a human panel.

Understanding Results

Model Outputs

Reports may include:

archetype count or mathematical iteration count
response distributions inside the simulation
segment or archetype breakdowns
model-generated themes and suggested follow-up questions
model-simulation ranges or other internal diagnostics

An iteration count is not a count of people. An internal range is not a classical confidence interval for a human population.

Reference Comparisons

When a relevant published statistic is available, selected model responses may be shown beside it. This is a diagnostic comparison and does not establish predictive accuracy. See Calibration.

Using the Output

Use survey-style simulations to:

narrow a list of messages or concepts
identify assumptions to test
prepare a human research guide
anticipate possible objections
decide what evidence to collect next

Validate important findings with real customers, a suitably recruited human survey, or live-market outcomes.

Exporting Results

Supported workflows may export model-generated results in formats such as PDF, CSV, or JSON. Keep evidence labels and limitations intact when sharing an export.

Products - Setting up products for surveys
Calibration - Understanding reference comparisons
Model Diagnostics and Statistical Methods