Estimating Software Costs: Bringing Realism to Estimating

Very thorough coverage on all aspects of software cost estimating. This book gives a compelling overview of all aspects of estimating software, to include very good coverage of maintenance and enhancement estimating. I have purchased other books by Capers Jones and I have to say this one is one of his best.

This second edition of the groundbreaking Estimating Software Costs is a must-have desk reference for every PM, program manager and estimation practitioner. It has been updated to include treatment of Agile methods (and other variants such as XP), OO development, UML, and CMMI. While IFPUG Function Points and Lines of Code are the still the leading measures of software size, Jones does provide data with respect to many of the emerging measures (story points, use case points, web object points, etc.) The book is divided into six sections. Section 1 presents a basic introduction to software estimation, including a brief history, capability and value of commercial estimation tools. There is also a very nice discussion on the potential sources of estimation error. Section 2 provides methods for generating early estimates and the danger that these will become accepted as THE estimate for the remainder of the project. Jones provides many simple rules of thumb for both classic size measures (Function Points and LOC) and emerging methods. Section 3 talks about methods of measuring size of various software work products. Again, the predominate method discussed is IFPUG Function Points; however, Jones does address the more abstract and "experimental" size measure in use today. Section 4 deals with the seven classes of influencing factors that drive project outcomes and how commercial estimation tools compensate for them. Jones concludes that industry averages for these factors should be discarded in favor of specific values from the performing organization. This reduces uncertainty and the political impacts. Section 5 defines ten activities that are common to many projects for the purpose of accurately deriving a bottom-up estimate. The implication of each of these activities with respect to software estimation is explored in detail. Section 6 examines the difficulty of maintenance estimation based on the notion of "software entropy," which is analogous to the Thermodynamics property of isolated systems. Entropy is a measure of disorder in an isolated system and increases with time. As a product ages, its level of disorder increases due to the number of maintenance patches and enhancements applied. This reduces the maintainability of the product and increases the difficultly in maintenance estimation. Again, Capers Jones proves to be a master at collecting, interpreting and presenting useful data. While some of the material (notably the rules of thumb) may be slightly over-approximated to be useful, Jones does present many ways to develop the initial early estimate and start the open dialogue that will ultimately lead to a successful project.

This book contains lots of useful data on the software development process -- including productivity -- using both LOC and IFPUG Function Points. It includes basic rules of thumb for estimating schedules and effort. It is a good starting point for organizations that do not have their own historical data to base their estimates on. An excellent reference to have within reach if you work with software measurement.

This is an excelent textbook about software projects estimation. The only problem with the book is the lack of more examples to learn the techniques discussed by Jones.

The software industry has an abysmal record of slipped schedules and cost overruns. With their fixed deadline, multimillion-dollar Y2K mitigation projects--the most complex in history--push the envelope of uncertainty and disaster. Estimating Software Costs, just released from McGraw-Hill, should help. Written by T. Capers Jones, the pioneer and leading authority in the field, this 724-page book is the definitive encyclopedic reference and a must-read for every Y2K and IT project manager. It covers all aspects of the problem, including commercial software tools, handling excessive schedule pressure, international factors and contractual and legal concerns. Jones' lucid writing style makes this otherwise leaden subject come alive. END