Type safety is the main industry-oriented selling point for languages like Haskell or Rust. It promises fearless refactoring, confident pushes to production, and nimbly pivoting to ever-changing business needs or market pressures. However, anyone who has taken a cursory glance at how many terrorist attacks the TSA has actually stopped has a deep appreciation of the high opportunity cost of security theater. At airports, it's long lines, wasted square footage, and spent tax dollars for no benefit. With programming, it's heavy-weight-yet-leaky-abstractions, overwrought CI processes filled with flaky, error-prone tests, or release approval ceremonies overburdened with theater and ritual. This talk will tackle one such example, using Yesod and Servant as examples in a case study of libraries that took radically different approaches to solve a near-identical problem, and what lessons engineers can take from them to improve their craft. Doing so will, Adam believes, enable them to write better, more easily changeable architectures, which will enable their department to be a source of strength to the firms that employ them, rather than a home of friction and delays.