IT/business alignment is one of the main topics of information systems research. If IT artifacts and business-related artifacts are coupled point-to-point, however, complex architectures become unmanageable over time. In computer science, concepts like the ANSI/SPARC three-level database architecture propose an architecture layer which decouples external views on data and the implementation view of data. In this paper, a similar approach for IT/business alignment is proposed. The proposed alignment architecture is populated by enterprise services as elementary artifacts. Enterprise services link software components and process activities. They are aggregated into applications and subsequently into domains for planning/design and communication purposes. Most design approaches for the construction of enterprise services, applications and domains are top-down, i. e. they decompose complex artifacts on a stepwise basis. As an alternative which takes into account coupling semantics, we propose a bottom-up approach which is demonstrated for the identification of domains. Our approach is evaluated using a telecommunications equipment case study.
Organizations constantly adapt their Information Systems (IS) architecture to reflect changes in their environment. In general, such adaptations steadily increase the complexity of their IS architecture, thereby negatively impacting IS efficiency and IS flexibility. Based on a Complex Adaptive Systems (CAS) perspective, we present a more differentiated analysis of the impact of IS architecture complexity. We hypothesize the relation between IS architecture complexity on the one hand, and IS efficiency and IS flexibility on the other hand to be mediated by evolutionary and revolutionary IS change. Subsequently, we test our hypotheses through a partial least squares (PLS) approach to structural equation modelling (SEM) based on survey data from 185 respondents. We find that the direct negative impact of IS architecture complexity on IS efficiency and IS flexibility is no longer statistically relevant when also considering the mediating effects of revolutionary and evolutionary IS change.
Enterprise architecture management (EAM) has long been propagated in research and practice as an approach for keeping local information systems projects in line with enterprise-wide, long-term objectives. EAM literature predominantly promotes strictly governed and centralized coordination mechanisms to achieve the promised alignment contributions. Notwithstanding the increasing maturity levels in practice, organizations still struggle with the successful establishment of EAM, mainly due to the inherent challenges of a firmly centralized approach in complex organizational settings. This study opts for cooperative learning as a theoretical lens to afford a distinctive, non-centralized conceptualization of EAM. We empirically demonstrate EAM as a stage-wise learning process in which knowledge acquisition and cooperative interactions among individuals contribute to project performance on the local level. Projects that benefit from this particular learning process, in turn, are found to significantly leverage enterprise-wide performance.