Episode Overview

Andrew Cooper, founder and CEO of Verve, on synthetic personas, the respondent crisis, and a new model for AI-powered insight.

Episode Highlights

• From access panels to AI simulations: Andrew traces Verve’s evolution from one of the early online panel companies to a specialist in AI-powered synthetic personas — built on the conviction that “a good simulation is better than an average sample.”
• Addressing the respondent crisis: Rather than replicating large samples artificially, Verve combines proprietary client data with curated sources to build auditable, validated simulations that correlate at 0.9 with real human responses.
• Where synthetic personas work best: B2B and specialist audiences — where recruiting real respondents is slow, expensive or unreliable — are proving to be the strongest use cases, alongside healthcare and financial services segments.
• Humans and simulations working together: Verve’s model keeps real people involved in the insight process, using simulations to handle onerous or tactical tasks, while human respondents contribute depth, validation and qualitative richness.
• The skills challenge for the industry: AI is reshaping the capabilities research businesses need — consultative skills, critical thinking and an AI-native mindset are growing in importance, while routine analytical work is declining.
• Oracles and internal AI agents: Verve has built 60 internal AI tools (“Oracles”) covering specialist areas of their workflow, effectively doubling productive capacity without adding headcount.

About the Guest

Andrew Cooper is the founder and CEO of Verve, a UK-based insight technology company specialising in AI-powered synthetic personas and simulations. He previously co-founded Research Now, one of the pioneering online access panel businesses.

With over two decades in the research industry, Cooper has been an early advocate for respondent quality and engagement, and now leads Verve’s work helping brands including Mars and Samsung generate board-level insight through a combination of real human data and intelligent simulation.