Abstract

Business outages are not new problems encountered by firms that rely on IT, but cloud computing poses new challenges. What makes cloud services unique compared to on-premise services is their shared-responsibility model: The Cloud Service Provider (CSP) and users share responsibility for maintaining the stability and availability of the cloud system. Service downtime can be caused by the CSP due to power outages, hardware and software bugs, etc., and by the user due to misconfigurations, human error, etc. Therefore, there are clear responsibilities and incentives for both parties to reduce service outages and to make investments (efforts) in cloud service availability. Based on this, new issues and challenges arise: First, what are the CSP’s and users’ optimal availability efforts, respectively? Do the efforts put forth by both sides inhibit or facilitate each other? Second, how do the exogenous market parameters affect the CSP’s and users’ incentives to exert efforts? Are the improvements in service quality effective in increasing the CSP’s profitability? Third, how is the CSP’s pricing decision influenced by the shared-responsibility model? To answer the above questions, we propose a game-theoretic model where the CSP aims to make optimal investment and pricing decisions to maximize its profit. Both the CSP and cloud users have incentives to exert efforts to improve service quality. We derive the CSP’s optimal pricing and investment decisions, as well as the optimal effort of users. This study reveals several interesting insights. First, we identify the market as being in low-loss and high-loss environments based on the loss of the unit downtime. Accordingly, the CSP exhibits different responses and makes different effort decisions under these two scenarios. We find that as the CSP’s responsibility ratio increases, the CSP may not always have an incentive to exert more effort. Second, considering the shared-responsibility feature of cloud services, we pay much attention to the interaction of efforts between the CSP and users. We are surprised to find that (1) the responsibility for reducing service downtime may at some point fall entirely on the users, which means the efforts of the CSP and users are sometimes not consistent, and (2) even if both parties put in efforts to improve quality, the CSP’s profits may still decline. Finally, regarding the pricing decision, the CSP will decrease the price only if its responsibility ratio is high enough as the CSP’s availability effort increases. This paper aims to (1) enhance cooperation between CSPs and users by emphasizing the shared-responsibility feature of cloud services, and (2) help CSPs make effective technology investments. This paper has strong management implications for the cloud services industry, and the findings provide valuable insights for the CSPs. Future research could lead to more interesting perspectives by incorporating the diversity of cloud service types and the fierce competition among multiple CSPs.

Download

Share

COinS