The majority of the book is broken out into chapters which each cover different types of estimation methods. Each of the chapters includes a chart at the beginning which summarizes the methods in the chapter, their potential accuracy, as well as time during a project when it can be used. As all the estimation methods are discussed, one important theme rings true throughout all of them. Collect historical data. Any data will help improve accuracy, even if its not as detailed as you'd like. Anything is better than nothing, and your company's historical data will always be far and away better than any "industry standard" data you may find on the Internet. Wherever possible, McConnell backs up his real world experience and advice with statistics from studies and clients taken over the years. While often times this sort of dry data obscures the real valuable information, the graphs and data presented in this book are formatted and explained very well, and add real value to each of the lessons being given.
As the new year is fast approaching, many software companies will be looking to create budgets for their upcoming projects and will be looking to you for good estimates. Do yourself a favor and read this book. Then present this as a gift to your manager (along with a copy of The Mythical Man Month) for Christmas. It will be well worth it.
Article comments