What is it?

A synthetic persona or AI persona is an AI-generated representation that mimics the responses and attitudes of real people.

Not to be confused with: digital twins

A synthetic persona is a representation of a consumer segment based on a generalised data set, while a digital twin is an AI simulation of one specific person, based on first-party data.

Why does this matter?

The language around synthetic data can be confusing, and there is a distinction between synthetic data and synthetic personas.

In a market research context, synthetic data refers to AI-generated responses that replicate the statistical patterns and characteristics of real data. Synthetic personas refers to AI-generated models that generate observations that reflect how a target segment is likely to respond, based on patterns from data sets.

There is a need to understand what synthetic personas are – and what they are not – because their use in commercial research is increasing.

What does it mean for research?

Brands and researchers can use synthetic personas for early-stage, exploratory research, such as evaluating product concepts or testing survey designs. Organisations are increasingly combining synthetic personas with human panel research.

Sam Barton, head of insight & effectiveness at Bountiful Cow, and co-author of a 2025 paper on synthetic data published by MRS, said: “Synthetic personas offer the chance to explore at speed and scale. By modelling how a group, or groups, are likely to respond, they can help researchers interrogate hypotheses, test concepts, and refine real‑world studies before going into field. Once created, they can remove the resource barriers that prevent some projects going ahead. They create commercial value, too: sharper questions, faster learning and richer outcomes.

“However, synthetic personas are not a replacement for human research. They work best as a complement to it, with live data and participant insight still essential for nuance, full validation, and more robust decision‑making.”

• Previous explainers in this series have looked at agentic AI and digital twins.