Failsafing service quality

Service quality matters, but variation based methods are of limited use in service settings because of violation of their underlying assumptions, and the preponderance of service failures arising from human error. There are more appropriate methods to control and decrease the likelihood and impact of human error, and failsafing is the best place to start, as it directly attacks the problem, is fast to implement and easy to understand and apply. There are three major steps to failsafing your service: identification of important errors, location of control points, and the creation of a failsafe device or procedure at the control point. We provide a framework for aiding the location of control points, and a taxonomy as an aid in the development of failsafe devices.