Engineering Systems: Best-in-Class/Worst-in-Class
Journal of Cost Analysis and Parametrics
Measurements goal is to help assess performance to determine which methods are productive or counterproductive. Metrics are tools used to identify and implement practices that lower costs, reduce time to market, and improve product quality. But process improvement is not accomplished through measurement or metrics alone. Rather, one must use the data to make conscious decisions that change the way business is done. In fact, one of best ways to make those decisions is by studying the characteristics of best- and worst-in-class software projects. Referencing Quantitative Software Managements database of 10,000+ completed software projects, this article evaluates the common factors that define the most and least successful engineering projects drawn from the databases System Software, Scientific, Telecom, and Command and Control application domains. Presenting a thorough analysis of project staffing, effort, duration, cost, and quality data, this article gives project managers a solid, scientific framework for evaluating potential projects and identifying winning strategies.
Donald M. Beckett is a principal consultant at Quantitative Software Management. He has over 30 years of experience in software and has focused on software measurement and estimation since 1995. He previously published a study on Best-in-Class / Worst-in-Class business IT projects in The QSM Software Almanac, Application Development Series.