Factor reduction and clustering for operational risk in software development

Abstract

Software development failures frequently emerge as a result of the failure to understand and identify risks.The aim of this paper is to identify the most salient risk factors during a software development project lifecycle, in terms of occurrence likelihood and impacts on cost overrun. A questionnaire survey was circulated to 2000 software development companies, IT consultancy and management companies, and web development companies in the UK, USA, Europe, India, China, Japan, Canada, Australia and Asian countries.This asked respondents to evaluate a number of risk factors. However, many factors were closely related and so we apply a factor reduction and clustering process to allow a smaller number of crucial risk factors to be identified. The three main clusters of risk factors identified in this study are 'feasibility study', 'project team management', and 'technology requirements'. While 'feasibility study' may be unlikely to occur it can have significant impact on outcomes; 'project team management' is likely to occur but has relatively little impact on outcomes in comparison to 'technology requirements'. Professionals will need to carefully check and balance these factors and generate a risk mitigation plan to reduce the severity of the project failures. These results allow them to connect the probability of occurrence and overall impact to focus their scarce resources on reducing the most pertinent risks in their project.Keywords: factor reduction, factor clustering, software development risk, risk occurrence likelihood, cost overrun INTRODUCTIONSoftware projects are dynamic and tend to have volatile requirements making them difficult to manage and control as the project scope can change frequently. Failure to understand and to identify risks leads to software development failures (Stahl & Bosch, 2014;Reyes et al., 2011). Literature on how to manage a development project often refers to cost, time, and quality as the key project success criteria; however, there are also many other different, broad, and overlapping definitions of project success and failures (Portillo-Rodríguez et al., 2014;Jorgensen, 2010;Baccarini et al., 2004;Linberg, 1999;Ropponen, 2000). Various researchers observe that in practice it may be very difficult to claim the project was really successful or a failure (Martín & Yelmo, 2014). A single project can be considered successful by one stakeholder and failure by another (Naquin & Tynan, 2003;Thüm et al., 2014). This is a copy of the "Post-print" (i.e., final authors' copy, post-refereeing). Published as:Mohd-Rahim, F. A., Wang, C., Boussabaine, H., . Factor reduction and clustering for operational risk in software development. Journal of Operational Risk, 9(3), 53-88. IT literature has produced a number of conceptual frameworks to explain different types of software development risk, risk management strategies and measures of software project performance (Dingsoyr et al., 2012;Nidumolu, 1996;. Many studies suggest that failure to manage risks causes common problems such as cost...