Short, plain definitions of the terms used in synthetic user research.
Silicon sampling is the practice of conditioning a language model on sociodemographic backstories so its outputs approximate the response patterns of human subgroups.
Synthetic data in market research covers two different things: respondents simulated by a language model, and statistically similar copies of a dataset you already hold.
A synthetic persona is a simulated research participant: a language model conditioned on a demographic and psychographic profile so it can answer in character.
A synthetic respondent is an AI-simulated survey participant — one row in a quantitative dataset generated by a language model rather than answered by a person.
Synthetic users are AI-generated research participants: language-model personas conditioned on a described population and asked the questions you would put to a real panel.
