The public cloud promises agility, yet we document significant customer inertia in a "best-case" scenario for switching: moving between products from the same provider, at the same data center, and within the same product series. Using proprietary customer-level data, we isolate a setting in which conventional switching frictions—such as migration fees and engineering costs—are minimal, allowing us to establish a lower bound on inertia in the public cloud market. Our findings suggest that this inertia is driven primarily by non-pecuniary organizational frictions rather than provider-imposed constraints; thus, even if regulators banned egress fees or standardized APIs, cloud customers would likely remain sticky. We develop and estimate a novel multiple discrete-continuous demand model to quantify this inertia. Surprisingly, we find that the cloud provider itself is also significantly harmed by this inertia, losing 58% of potential revenue as customers fail to adopt superior technologies that would drive usage expansion. While customers still derive substantial surplus from the cloud, inertia reduces total welfare by 62%.