Researchers design a pricing system for data markets that learns what buyers will actually pay

A team of economists built a two-stage auction system that first tests the market with traditional auctions to see what data is worth, then switches to fixed prices based on what it learned. In practice, this means data sellers can now set prices that adapt to what buyers actually reveal about their willingness to pay, rather than guessing or using fixed price lists.

Data marketplaces are growing fast, but pricing data is hard because every buyer values it differently and sellers have almost no information about demand. This system solves that by making the learning public and measurable: instead of setting prices blindly, a seller runs a small auction first to understand the market, then locks in prices that work. The practical effect is that smaller data sellers who can't afford sophisticated pricing strategies suddenly have access to a method that large firms use.