1

Abstract: Several organizations have institutionalized comprehensive metrics programs to strengthen management decision making and enable continuous improvement of software engineering processes. A range of software metrics focusing on various quality attributes and abstraction levels have been recommended by the research community. Recent studies indicate, however, that almost 80 % of the software metrics initiatives do not succeed. The study depicted in this paper is a reflection of the current state of adoption of software metrics in industry vis-à-vis research going in academia and highlights areas for further improvement and future research. The journey ahead for software metrics lies in using relatively simple existing metrics to formulate management decision-support systems that associate different aspects of software development and testing processes and equip managers to make many types of evaluations, predictions and trade-offs during the entire software life-cycle. The approach suggested in literature is to handle the key considerations largely overlooked in the usual metrics models namely: causality, ambiguity and different associated subjective issues. Therefore, the way ahead for research on software metrics lies in causal modelling, empirical software engineering, and multi-measure decision aids. This paper is an outcome of a thorough literature review exercise carried out on manuscripts, articles and findings related to software metrics presented over past four plus decades accompanied by the experience and observation of authors in the domains of industry and academia. Th