You talk to customers, you collect data, you got buy-in from your internal stakeholders. You lined everything up for a successful feature launch. And then you run an experiment, and the reality showed you that the implementation that you expected to be the winning one is actually not what moves the needle for the larger product. Now you have a dilemma - you have the option to launch as is (after all, you did the due diligence - maybe the experiment timing was bad and in the real world, people will love what you did) or go back to the drawing board.
You are a PM. You work for a big company or a startup. You are tasked with determining the direction for your product, so you enthusiastically embark on the journey to talk to as many customers (or potential customers) as possible to determine what their needs are and how your product can help them in the long-run. After months of painstakingly collecting feedback, you emerged victorious with a PowerPoint deck that says: “We need a bigger button in the toolbar to let people renew their subscription easier”.