Rhys Fisher is the COO & Co-Founder of Rally. He previously co-founded a boutique analytics agency called Unvanity, crossed the Pyrenees coast-to-coast via paraglider, and now watches virtual crowds respond to memes. Follow him on Twitter @virtual_rf
In a battle of AI decision-making, we put straightforward GPT predictions head-to-head with virtual audience simulations. Can basic GPT models, relying on pure heuristics, match up to a synthetic audience that mirrors human nuance? In this experiment, we challenged GPT 4o, o3-mini, and a role-prompted GPT 4o-mini against Rally’s tailored virtual audience.
With synthetic research and persona-driven prompting still in beta, little is known about how different AI models shape responses. In our simulated buying committee experiment, we found that more powerful models can replicate the complex decision-making seen in high-stakes human scenarios. If you want your synthetic research to truly mirror human thought, here's how choosing between Fast and Smart mode can unlock insights that reflect real-world nuance.
Enterprise sales can feel like herding cats—with each stakeholder more impossible than the last. Nail it, and you’re hitting yearly targets; mess it up, and you’re bankrupt in epic style. So I set up the most absurd experiment ever: I built a fictional buying committee in hard mode, then got them all to say, “take our money!”
The OKCupid study showed us that racial preference in dating is real—and stubborn. But what if you could run your own experiment, and ask 400 virtual women exactly what would make them swipe right… even if you’re not their usual type?
Even basic personas are useful for gathering feedback. Since they role-play better with rich context, you can leverage that to build the business case for further research.
