Objective Based Performance Measurement

Paper Abstract Objective metrics-based evaluation of any project and person is essential for management. As put in succinctly by a renowned manager that if you cannot measure 'it' you cannot manage 'it'. Evaluation of software projects and software engineers has been an elusive target for software engineering practitioners. How to say that one software engineer is better compared to another and in which kinds of projects is a key question, yet to be answered satisfactorily. As software increasingly pervades every sphere of life, there is an increasing pressure on industry to deliver better software in smaller span of time. This makes answering of such questions essential as any industry can mature and standardize on processes and metrics only. The authors developed a model for objectives-based performance measurement (OBPM) at their company, which is into development of software products. This OBPM model uses a series of metrics categorized into three houses of "efficiency", "quality" and "logic". Efficiency is about writing code quickly, quality is about writing error-free code and logic is about writing resource-efficient and good code. The metrics used are objective, with some evaluated directly based on inputs while others are based on feedback from users of the code and those who observe the usage of the code. OBPM uses mean and variance, as the major statistical measures, as they have been proved to be very effective in statistically describing a process and are used in Six Sigma quality model also. OBPM also uses "Function Point" standard, an established concept in software industry, of measuring a software project's complexity and effort, as the baseline measure. For the past six months, OBPM is being used at the company where the authors are working. It was discovered that OBPM model was able to show patterns and inferences, which were otherwise found to be true and realistic. The model also helped the software engineers know their performance and areas where they can improve. This resulted in a better-focused training. Normally training of engineers is restricted to common ones on technologies and soft skills. OBPM model is able to identify exactly where an individual needs to be trained. OBPM m odel can also be used to do predictive analysis. It can help determine which engineers would be most suitable for a crunch project where meeting the deadline is absolutely essential. The paper also discusses some of these findings and results. The authors plan to …