Mike Kelly builds on Scott Barber's work to show how you can combine performance-degradation curves and complex performance scenarios to help determine "good enough" quality for an application in terms of performance. Read the article here.