Slide 9 of 61
Notes:
One of the major problems with cost overruns is that many methodologies or project plans call for an unreasonably precise estimate of costs before the project begins. These estimates are made after a quick preliminary study or feasibility study. Think about it! Can you accurately estimate project costs before making a detailed study of the current system or defining end-user requirements? Can you estimate the costs of computer programming before a detailed systems design has been completed? It's not very likely. The cost estimates of a project will change as you get further into the systems development process.
Poor estimating techniques are another cause of cost overruns. We suspect that many systems analysts estimate by making the best calculated estimate (guesstimate?) and then doubling that number. This is hardly a scientific approach.
Cost overruns are often caused by schedule delays. Once again, we can point to premature estimates as a problem. These early estimates are based on the initial scope of the project. Because systems analysts (and information systems professionals in general) are eternal optimists, they often quote optimistic schedules and fail to modify those schedules as the true scope of the project becomes apparent.
Because many managers and analysts are often poor time managers, project schedules slip slowly but steadily. “So we've lost a day or two! It's no big deal. We can make it up later.” This may be true, but then again, it might not. They fail to recognize the fact that in the systems development life cycle certain tasks are dependent on other tasks. Because of these dependencies, a one-day slip can set the whole schedule back. And when those one-day delays pile up, we inevitably find ourselves working 15-hour days at the end of the project.
Another cause of missed schedules is what Brooks (1975) has described as the mythical man-month. As the project gets behind schedule, the project leaders frequently try to solve the problem by assigning more people to the project team. It just doesn't work! There is no linear relationship between time and number of personnel. The addition of personnel creates more communications and political interfaces. The result? The project gets even further behind schedule.
