The tool, called YouGov Parallax, takes its name from the optical principle of parallax, whereby two different viewpoints, when combined, can create three-dimensional depth.

YouGov Parallax uses individually mapped AI twins using data from real YouGov panel members to simulate responses across a wider range of questions instantly and at scale.

The second perspective offered by the tool is to run surveys to validate findings or deepen their understanding of consumer sentiment, which can be done using a subset of the panel members whose twins were used, a fresh general population sample, or a targeted audience of their choosing.

YouGov said that validation can be applied to the full question set or focused on the most critical questions in a given study, with sample sizes, turnaround times and the depth of validation all configurable. 

The company added that the simulated layer aims to provide breadth and speed, while the validation layer targets certainty and a calculable margin of error. 

Stephan Shakespeare, chief executive and co-founder at YouGov, said: “We’ve been sceptical about synthetic data, and we were right to be – used without proper grounding, it gives you convincing-looking answers that can be fatally flawed and cannot state reliable confidence levels.

“Two views, properly combined, give you depth that neither can give you alone.”