Resource flexibility, arguably among the most celebrated operational concepts, is known to provide firms facing demand uncertainty with such benefits as risk pooling, revenue-maximization optionality, and operational hedging. In this article we show that resource flexibility mitigates censoring of sales data, and thereby facilitates learning the true demand under a variety of practical settings, demand distributions, and model parameters. This suggests that extant models, which view flexibility only as the ability to act ex post, could be underestimating its true value when learning the demand is desirable, for example, when it enables managers to make better-informed decisions ex ante.