Any business trying to sell its products profitably must have some idea of what sales and hence revenue would be at different prices. Suppose we double our prices. Sales will fall, but will margins go up so much that it is worth it? Suppose we halve our prices. Margins will fall, but will sales go up so much that it is worth it? In making these judgments, a business might use historical data or simply guess by adding a mark-up over costs. But there is a better way to learn the structure of demand, one that uses real time information and discovers the revenue-maximizing price at the same time. It is a variant on the so-called Dutch auction that has been used for over a hundred years in the Netherlands to sell flowers. In a Dutch auction, the price starts off high and is lowered until someone buys. Dutch auctions are fast: only one person has to say they want the object for sale and the auction is over. We used a variant of the Dutch auction to set prices for basketball games at Northwestern University where we work at the Kellogg School of Management and the Department of Economics. We call it Purple Pricing. Our Dutch auction has one vital and important twist: the Purple Pledge guarantees that, wherever the final price ended up, if someone buys a ticket at a higher price, they get a refund for the difference between the price they paid and the final price. The Purple Pledge is the key feature that allows us to learn demand and set the optimal price at the same time. Get the full story at Harvard Business Review or listen to the interview at the HBR Blog