Users’ Loyalty to Agile Information Systems

Over the past few years, across many industrial sectors, Information Systems (IS) developed with the help of agile methods have become the rule rather than the exception. Because of their high flexibility, such Agile IS development methodologies help firms to keep pace with emerging market requirements. At the same time, customers are also gaining increasing market power due to an expanding digitalization of services and products, which decreases switching barriers and increases transparency. As a result, it has become crucial for firms to develop IS that continuously provide sufficient value to customers. This is one of the main reasons why firms regularly deliver increments of Agile IS for users to update outdated software versions. By doing so, firms try to bind and engage customers lastingly to capture current and future revenue streams and stay competitive. Agile IS and software updates (that deliver increments of Agile IS to users) have been researched thoroughly, however mostly from a technical point of view. Nevertheless, because updates change a system while it is already in use, they have the potential to impact users’ beliefs, attitudes, behaviors, and in particular, loyalty to a software in the post-adoption phase. However, despite the importance of better understanding user responses to Agile IS to provide an adequate theoretical framework, research from a user’s perspective on Agile IS, and especially software updates, is still scarce. Against this backdrop, this thesis presents four empirical studies that were conducted to investigate whether and how Agile IS affect users’ loyalty to IS, to identify potential moderators, and to understand how Agile IS should be designed to facilitate potential positive effects. In these studies, increments of Agile IS are operationalized as software updates and customer loyalty as a user’s continuance intention with a system. By drawing on the IS Continuance Model in a scenario-based online experiment, the first two studies reveal empirically how Agile IS have the potential to increase user continuance intentions. Users of Agile IS show greater IS continuance intentions, despite that some functionality is provided only later on, as compared to a consistently feature-complete traditional IS. This effect is diminished somewhat when the software is introduced with an extensive feature set right from the beginning. Nevertheless, the size of an update does not seem to play a significant role. The second study reveals that this positive effect of updates only emerges if the user is not very knowledgeable regarding the software, because experts in contrast to novices seem to devalue Agile IS (their continuance intentions decrease with Agile IS in comparison to traditional IS). Additionally, the second study shows that the removal of features through updates reduces continuance intentions even more than the equivalent addition of features when considering the absolute magnitude of change. With empirical data from a laboratory experiment, the third study identifies update frequency and update type as further moderators of the effect, and confirms the hypothesized mediation mechanism presumed by the IS Continuance Model. The fourth study examines the role of update delivery strategies, i.e., the timing and presence of a notification and an installation choice. In this study, feature and security updates are distinguished, as both seem to have different characteristics with respect to the delivery strategy (i.e., users ‘need’ security but ‘want’ to add functionality). The findings show that both update types should be announced to users, in the case of a security update, only after successful installation, while presenting an installation choice to users prevents any positive effect for all types of updates. Overall, this thesis highlights the importance of understanding Agile IS and software updates from the user’s perspective. First, the results show that Agile IS have the potential to affect user’s continuance intentions, thereby contributing to a comprehensive theoretical foundation on Agile IS. Also the findings put the user more at the center of investigations in IS. Second, the empirical findings provide evidence in support of a necessary fine-grained understanding of IT Artifacts as malleable compositions of specific features and characteristics. This answers the call of several researchers to put the IT Artifact more at the focus of IS research (Benbasat and Zmud 2003). Third, the results reveal that changes in IS might change users’ attitudes and behaviors over time, which extends the predominant view of IS in post-adoption literature from a mostly static to a more dynamic perspective. With this finding, we answer the call of several IS scholars to consider the evolution of IS more thoroughly (e.g., Jasperson et al. 2005; Benbasat and Barki 2007). For practitioners, the findings of this thesis provide empirically backed rationales to inform management decisions concerning the deployment of Agile IS and offer guidance on strategic or design considerations. Overall, the results show how and when the value provided by IS from a user’s perspective may be increased by the deployment of Agile IS and software updates.