What Drives Application Portfolio Complexity? An Empirical Analysis of Application Portfolio Cost Drivers at a Global Automotive Company

In today's economy organizations have to continuously adapt to changing external requirements. This often necessitates the adoption of new technologies and consequently the implementation of new applications. As a result, enterprise architects struggle with increasingly complex application portfolios that are costly to maintain and operate. In order to adequately manage these application portfolios, a better understanding of the relation between business application characteristics and operation costs is required. In this research we therefore analyze data from 3656 business applications of a global automotive company to evaluate the effects of commonly conjectured application portfolio complexity drivers. Our results show that several application characteristics-in particular the number of supported interfaces, business processes, business data, and users-correlate with operation costs and with incident numbers. We furthermore find that these correlations differ significantly for different application types.