Electric Twin

The company explains its accuracy metrics for synthetic audiences: 1-MAE, NDAM, and hold-out testing.

Synthetic audiences built from policyholder data enable testing of pricing, products, and claims communications.

Electric Twin allows testing of account fees, product propositions, and customer communications using survey-based synthetic audiences.

Synthetic audiences from subscriber data help assess reactions to pricing, content, and format decisions.

The Electric Twin team provides research notes on synthetic audiences, AI market research, and testing messaging with customers.

Electric Twin’s synthetic audiences help enterprises like *The Times* and *Virgin Red* scale consumer research and validate decisions.

Ben Warner is a data scientist and co-founder of Electric Twin.

Electric Twin’s research agent answers questions by designing an approach, running research, and generating a shareable report.

Electric Twin’s press coverage includes mentions in *The Times*, a $14 million funding round, and the *Entrepreneur UK 100* list.

Electric Twin lists open roles for its team building synthetic audiences for enterprise research.

Electric Twin compares its synthetic audiences to Claude and ChatGPT, highlighting superior performance for consumer insights.

Michael Muthukrishna is Science Advisor to Electric Twin and Professor of Economic Psychology at the London School of Economics and NYU.

Electric Twin demonstrates how synthetic audiences provide fast consumer research answers using customer data.

Answers to common questions about Electric Twin's synthetic audiences: accuracy, how they're built, data security, pricing, and vs ChatGPT.

Electric Twin builds synthetic audiences from customer data, delivering 96% accuracy (1-MAE) and LSE validation.