Glossary (starter set)
Clear definitions designed to reduce category confusion.
Synthetic market research
Research workflows (surveys, concept tests, message tests, scenario simulation) run using statistically grounded
synthetic panels. Best treated as an accelerated way to generate and test hypotheses, then validate with benchmarks
and, where appropriate, targeted fieldwork.
definition
validation
disclosure
Synthetic personas
Simulated individuals with attributes and constraints (demographics, behaviours, preferences) designed to represent
segments of a target population. Credible systems are explicit about what grounds the persona and what is inferred.
panels
grounding
limitations
Digital twins (in research)
A stricter term often implying higher-fidelity, more stable, and more carefully grounded representations that can be
tested repeatedly. In practice, the label varies - so buyers should ask for validation and disclosure, not terminology.
stability
reporting
Synthetic respondents / synthetic samples
Synthetic “participants” used to answer survey-style questions. The key risk is confusing plausible outputs for measured reality.
Require population framing, repeatability, and external benchmarks.
surveys
safeguards
Want the full taxonomy?
The full glossary includes comparisons (synthetic data vs synthetic research, personas vs digital twins),
plus definitions optimised for consistent procurement language.
Buyer guide
A practical path for evaluating tools, agencies, and vendors.
Templates
Ready-to-use documents for consistent studies and disclosure.