Service-Level Agreement Negotiation in Cloud Computing Buying Organizations

Introduction

Cloud computing (CC) is characterized by on-demand readiness of computing re-sources, like storage, processing power, and networking speed, without the direct interaction of the user (Schneider & Sunyaev, 2016). However, the technology service is often summarized and agreed upon by the Service Level Agreement (SLA) that are not fully capturing the nature of the relationship and responsibilities between the vendor and the buyer, causing complex and ambiguous agreements that are leading to trust issues and misunderstanding of the technology itself. Consequently, additional internal transactional costs are necessary to fully enable the service, leading to cheaper outsourcing strategies in order to prioritize external transactional costs. The SLA is “an agreement between the support service and the user, quantifying the minimum acceptable service to the user” (Hiles, 2000, p.11). A cloud SLA, is an agreement between a Cloud Service Provider (CSP) and a Cloud Service Customer (CSC). In this study, when mentioning SLA, it is always in the context of CC. In the majority of the cases, SLA is established between a seller and a buyer organization. However, SLAs may be agreed between two or more companies (Ibrahim, Kliazovich, & Bouvry, 2016). The increasing growth of agreements with CSPs and other third-party vendors is driving more pressures towards IT to mature comprehensive SLA to facilitate the governing of the IT services signed for and to guarantee that maximum value is conveyed to the CSC (Al-Gharibi, Warren, & Yeoh, 2020). This makes it imperative for both the CSP and CSC to priorities a correct SLA management, enforcing and encouraging cloud SLA standards (Girs, Sentilles, Asadollah, & Mubeen, 2020) that not only promise consistent SLA negotiation but also ensuring flowing communications among all engaged parties in the service delivery process and delivery stay open and dynamic. Both CSP and CSC require a way to associate SLA performance to the business values they are supposedly committing to. This research objective is to highlight this void with the Transaction Cost Theory (TCT) (Williamson, 1985). The TCT underlines that within exchange relationships, there is a fundamental deficiency of trust, followed by significant issues like uncertainties and opportunisms that surround transactions (Sutcliffe & Zaheer, 1998; Zhong & Myers, 2016; Ayaburi, Maasberg, & Lee, 2020). These notions, throughout the analysis of the Cloud SLA as a pivotal research element, may be of significance in analyzing organizations' willingness to accept and implement CC services. An understandable cloud SLA is necessary to solve the issues involved in the dependability, reliability, and security of the cloud for each party (Undheim, Chilwan, & Heegaard, 2011). The cloud SLA should be organized with satisfactory clauses and details to reach the user expectations, but also that SLA should be effortlessly evaluated to create accountability in case of legal breaches (Dillon, Wu, & Chang, 2010). Furthermore, SLA is an essential aspect of the CC service, and companies have been advocating for more complete SLAs (Kuma & Pradhan, 2013; Cochran & Witman, 2011). The above findings are problematic in the signing of the SLA and, consequently, for the adoption of CC. In this study, with “Dynamic SLA Iteration,” we mean the change of CSC needs in requirements and deliverables and how the changes have to be adjusted in the CC SLA. Therefore, the research problem addresses the fact that a considerable proportion of service level agreements in CC is inefficient and/or misinterpreted. To address this research problem, the authors have raised the following research question: “How to improve SLA management with dynamic SLA iteration?”.