One reason not to do that is the sales and marketing lead time. Every time marketing changes something there’s a 2 week lag before they can measure the conversion impact. With your usage-based trial concept that lag time becomes indefinite.