Researchers built a retail simulator that actually predicts how people buy things

Computer scientists created a simulation of the entire retail process—from how sellers pitch products to how buyers decide to buy—that reproduces real-world shopping patterns like price sensitivity and demographic differences. This means companies can now test sales strategies in a controlled environment before deploying them in the real world, instead of guessing or running expensive experiments on actual customers.

Until now, retail simulations only modeled pieces of the buying process in isolation. Nobody could test how a change early in the sales pitch would ripple through to the actual purchase decision because the simulators didn't connect those stages together. This one does. That means a retailer can run hundreds of variations of a sales approach, see which actually converts, and understand why—without burning money on live experiments that might fail.