Using data from more than 80 development projects, this paper attempts to answer the question: how much effort should be invested into code quality? It is shown that a "quick and dirty" approach is actually preferable in some situations. Volatility of requirements, expected breadth of usage, customers' defect tolerance, cost of defect fixing and system lifespan are suggested as the main factors determining how much effort to spend on improving code. These factors are then used to identify situations where XP coding practices are inefficient and to find boundaries within which simpler, less expensive methods give better results.