Grid computing can meet computational demands and offers a promising resource utilization approach. However, little research details the drivers of and obstacles to adoption of this technology. Institutional and organizational capability theory suggests an adoption model that accounts for inter- and intra-organizational influences. An empirical study with 233 high-ranking IT executives reveals that adoption results from social contagion, while organizational capabilities such as trust, firm innovativeness, tendency to outsource, and IT department size, influence adoption from an intra-organizational perspective. The findings show that mimetic pressures and trust play major roles in adoption processes, which differentiates grid computing from other inter-organizational systems.
In the past decade, a networked economy has evolved in which organizations collaborate and create supply chains or value networks. Such networks constitute relationship webs that generate both tangible and intangible value through complex, dynamic exchanges across organizations. The adoption of new technologies in such closely collaborating, networked economies can be consequential. Technologies such as electronic data interchange, voice-over-Internet protocols, and electronic business-to-business marketplaces, constitute a category of such technologies, referred to as inter-organizational systems (IOS) (Eom, 2005). Though the aforementioned have received pertinent research attention, another IOS gaining prominence has not – ‘Grid computing’. This study aims to address this gap by analyzing the influences on the adoption of grid computing.
Specifically, grid computing connects various IT resources through a physical network, offering members of the network the capability to share their IT resources (Foster et al., 2001), potentially improving utilization of otherwise idle resources. Grid computing thus can provide significant advantages to its adopters, especially for organizations that comprise subsidiaries in different time zones (where off-peak resources in one zone can be utilized on-peak in another zone). The concept of cloud computing evolved out of grid computing and often uses a grid as its backbone. This evolution reflects a shift in focus, from an infrastructure that delivers storage and computing resources (i.e., grids), to one that is economy-based and aims to deliver more abstract resources and services (i.e. in clouds) (Foster et al., 2008). Grid and cloud computing both can be employed internally by a firm, or can be exposed to others as an IOS. Furthermore, both grid and cloud computing appear poised to induce paradigm shifts, similar to the shift that marked transition from mainframe to client–server architectures in the early 1980s (Bhardwaj et al., 2010). Yet we know little about prospective users’ intentions to participate in this paradigm shift.
On first consideration, grid computing appears similar to other IOS with regards to factors that influence its adoption. On closer scrutiny however, several important differences become apparent. We will discuss these differences in detail in Section 3 and motivate why a distinct adoption model is appropriate for grid computing. We propose a combined model including institutional pressures and factors that are critical for grid computing, highlighting the differences from other IOS adoption models. In Section 2, we introduce the technology underlying grid computing and its strategic impacts on the firm. We highlight differences between grid computing and other IOS, then offer some examples from business practice before discussing the theoretical background for our proposed adoption model. We describe our conceptual model in Section 5 and subject it to several pretests. Section 6 outlines the status quo of grid adoption in business practice and presents the model results based on data from a large field study involving 233 IT managers who have responsibility for IT budgets. Additionally we compare the results of our adoption model with those obtained for other IOS. Finally, Section 7 concludes with a discussion of the implications of our findings and study limitations and future directions.
