Tom Mochal, along with many IT management pundits, have it only half right. Not only does the development effort need a clear understanding of the business requirements (an interactive learning process), but IT management needs a clear understanding of your development process, criteria for evaluating the quality of designs proposed to address the corresponding system/software requirements, and established mechanisms to calibrate design parameters (both scope and substance) to effort estimations. Without this latter capability, you are planning and scheduling without any basis in reality, which is the *real* cause of project schedule and cost over-runs. This approach is perpetuated by the “generalist manager” myth… that a good manager can manage any kind of project regardless of the scope and depth of knowledge they have relating to the development process. Way too much planning (and project management) is based on superficial knowledge and political constraints.
