In attempting to implement a simple, computational lot-sizing procedure in a factory producing automobile parts, I met substantial resistance, even though the company is renowned for using high technology in its products. I recommended a multiple-item economic order quantity model but was prevented from implementing it by interdepartmental politics. I realized that people can resist the force of logic, that dividing a plant into functional departments can impede progress, that people with little knowledge of well-established methods have little motivation to adopt them, and that even very small changes can yield large returns.