Where does the idea of a digital twin come from?
The idea comes from engineering: scenarios are tried on a virtual copy of a factory, an engine or a city, so the real system is never put at risk. Applied to consumers, scenarios such as a price change or a new package design are tried on virtual consumers first.
How is a digital twin of a consumer built?
At SCL, digital twins are not copies of individual real people. Profiles are built from statistical distributions and past consumer research, and each profile is represented by its own AI agent. The sample as a whole reflects the structure of the target audience, but no agent corresponds to a specific real person.
What is the difference between a digital twin, a synthetic consumer and a persona?
A marketing persona is a single representative character, and it usually lives on a slide. A synthetic consumer, or a digital twin of a consumer, is a working model that answers questions, and hundreds of them together form a sample. At SCL the two terms mean the same thing: a digital person with a defined profile who answers questions.
What are digital twins used for in market research?
Digital twins are used to answer “what if” questions without waiting for fieldwork.
- Price scenarios: how does preference change under inflation, a competitor’s promotion or a new product launch?
- Product decisions: which combination of features stands out in which segment?
- Communication tests: is the ad remembered, and is the message convincing?
- Focus groups: what are the reasons behind a choice?
What are the limits?
A digital twin is a model. It does not replace the real person. Its accuracy depends on the quality of the data behind it and on comparisons with real consumer data. The results point a decision in a direction, and the final decision still belongs to people.
| Marketing persona | Digital twin of a consumer | Real consumer panel | |
|---|---|---|---|
| What it is | One representative character | A model that answers questions | Real people |
| Scale | A handful of personas | Hundreds or thousands of profiles | Limited by budget |
| Testing scenarios | Not possible | Instant and repeatable | New fieldwork every time |
| Level of evidence | Assumption | Model output that needs validation | Direct statements |
Frequently asked questions
Is a digital twin the same as a synthetic consumer?
In consumer research, yes. Both describe an AI-based digital person that resembles a real consumer profile and answers questions according to that profile.
Are digital twins built from real people’s data?
No. Profiles are generated from statistical distributions and past research. They are not copies of a specific real person.
Can digital twins be built for Türkiye specifically?
Yes. SCL offers ready-made samples that cover all 81 cities and the socioeconomic structure of Türkiye’s population, or you can define the audience attribute by attribute.
How are the answers of digital twins validated?
Models are compared with real consumer data. The most robust approach is to read the synthetic result next to a small study with real consumers.
Sources
The assessments on this page draw on published academic work on simulating survey respondents with large language models. You can read our summaries on the articles page. Read the article summaries.
Related pages
- What is synthetic data?
- What is a synthetic survey?
- About SCL
- Example study: social proof and willingness to pay
Last updated: September 2026