One place I'd love to try prediction markets is with product changes undergoing A/B tests. A product manager could submit an experiment and people at the company could bet on what the product change will do to engagement/revenue/growth metrics. This framework could eventually be used to weed out product changes before they are built. It would be particularly useful when the product change is driven by the leadership and most of the rank and file think the direction is wrong.