#165 — How AI Is Killing Traditional Market Research: Peter Weinberg

Peter Weinberg is the founder of Evidenza, an AI-powered synthetic research platform, and a former LinkedIn executive where he co-founded the B2B Institute. Over a decade at LinkedIn, Peter helped reframe how B2B brands think about growth — shifting the industry's focus from bottom-of-funnel conversion toward brand building, mental availability, and reaching buyers before they enter the market. His work draws heavily on the Ehrenberg-Bass Institute's frameworks, and he's collaborated with some of the world's largest B2B organisations on marketing strategy and segmentation.

Most market research never gets done. It's too slow, too expensive, and the people you most need to reach — CFOs, in-house counsel, niche enterprise buyers — simply don't take surveys. So companies either skip research entirely or make decisions based on what the sales team heard from the three customers who called last week.
Synthetic research changes that equation. By using AI to simulate statistically representative populations of real customer types, organisations can now get directionally accurate, quantitative customer intelligence in a fraction of the time and cost of traditional research. Peter's company, Evidenza, has validated this approach across dozens of markets and categories — consistently finding 80–95% alignment between synthetic and human survey responses.

In this episode, we explore what that means for how companies understand customers, structure their innovation funnels, and rethink the long-standing political battle between marketing and sales over who really owns the voice of the customer.
In this episode we cover:

What synthetic research actually is — and why "lab-grown customers" may be more reliable than survey respondents clicking through for an Applebee's gift card
The accuracy question: how closely AI-simulated responses match real human data, and what it means when they diverge
How synthetic research could reshape the innovation funnel — moving from testing 3 ideas a year to testing thousands
Why the real opportunity isn't hyper-personalisation, but finding the mass-market commonalities that drive scale
The adoption barrier that has nothing to do with AI scepticism: organisations that wouldn't act on good market research even if you handed it to them

Episode Timeline:

• 00:00 — Highlight from today's episode
• 00:34 — Introducing Peter + the topic of today's episode
• 02:55 — If you really know me, you know that...
• 03:45 — What's your definition of strategy?
• 06:04 — Peter's decade at LinkedIn and the case for B2B brand building
• 07:41 — The Ehrenberg-Bass Institute: mental and physical availability
• 10:24 — What is synthetic research?
• 13:16 — Accuracy, speed, and cost: the three metrics that matter
• 14:54 — When human surveys lie (and synthetic respondents don't)
• 17:49 — Can you use synthetic research for internal adoption challenges?
• 19:59 — How synthetic research widens the innovation funnel
• 23:27 — Who owns the voice of the customer: marketing vs. sales
• 26:50 — Micro-segmentation vs. mass marketing — which does AI actually favour?
• 30:59 — Barriers to adoption: AI sceptics, soft rejectors, and market orientation
• 33:48 — When AI outperforms humans with AI (the doctor study)
• 35:57 — How to follow Peter and find Evidenza

Additional Resources:

LinkedIn: https://www.linkedin.com/in/weinbergpeter/
Website: evidenza.ai

Thank you to our executive producer Zach Ness, our producer Nazanin Homayoun Jam and our editor James Pearce. If you enjoyed this episode, please follow, download, and subscribe. I’m your host, Kaihan Krippendorff—thank you for listening.

Follow us at outthinker.com/podcast