CLOUD IMPLEMENTATION IN ORGANIZATIONS: CRITICAL SUCCESS FACTORS, CHALLENGES, AND IMPACTS ON THE IT FUNCTION

ABSTRACT

Organizations have been forced to rethink business models and restructure facilities through IT innovation as they have faced the challenges arising from globalization, mergers and acquisitions, big data, and the ever-changing demands of customers. Cloud computing has emerged as a new computing paradigm that has fundamentally shaped the business model in terms of how IT services are developed and delivered. A lot of conversations about cloud technologies can be observed in blogs, workshops, magazines, and industry reports, but there seems to be a paucity of either research or formal theorizing with respect to cloud computing implementation in organizations.

This dissertation develops a new theoretical framework with which to conceptualize the bidirectional relationships between IT departments and cloud computing implementation. The research seeks to achieve two key objectives. The first is to examine the factors that influence the success of cloud computing implementation in organizations. For this objective, a qualitative study using interview data and grounded theory was conducted. A taxonomy of both the Critical Success Factors (CSFs) and the challenges facing cloud implementation is developed, drawing on theories from organization and management studies. The study on CSFs and challenges indicates that a successful cloud computing implementation not only calls for a good understanding of cloud technologies, but requires a lot effort in, for example, mobilizing internal and external resources; dedicated architectural design; overcoming organization-specific challenges; and addressing political and cultural issues arising inside the IT department. The second objective is to investigate the impacts of implementing cloud computing on the IT function. To achieve this objective, a case study of two community cloud models was conducted. The study has identified a range of areas where the IT function can be impacted as a result of cloud implementation. These areas include the role and functionality; IT leadership; skills and jobs; IT staffing; formal structure; and workplace culture. Propositions with specific observations are provided to show how the IT function can be shaped in each affected area.

This dissertation contributes to the literature by extending extant CSF theory into the novel context of cloud computing implementation. It also develops new mid-range theories on ITenabled organizational change by exclusively focusing on the overlooked effects that an IT innovation can bring about within the IT departments. In addition, this research offers a number of actionable strategies for practitioners regarding how to make cloud implementation successful, and it highlights the importance of anticipating the potential business impacts of cloud computing on IT departments.

           

Chapter 1. INTRODUCTION

Globalization, mergers and acquisitions, tremendous amounts of electronic data, and everchanging customer demands have created a proliferation of intensive information-processing issues. These challenges require organizations to rethink their business models and structures, and these changes need to be facilitated by innovations in IT. In order to gain a competitive edge, organizations are forced to move forward to be more flexible, efficient, and innovative.

The emergence of cloud computing provides a promising solution for the challenges that confront modern enterprises. It is characterized by utility computing and web-based services, and while the technology is not new, the business model that it uses substantially impacts how IT resources and capabilities can be used to support business operations. Cloud computing is widely popular with small business startups because it allows new services to be created quickly without a large upfront investment, and it enables big firms to concentrate on their core business without worrying about scaling, failure, and the maintenance of computing capabilities. It has become a key strategy for IT vendors, information service providers, telecom service providers, and governments (e.g. Apps.gov in U.S. and ―The Kasumigaseki Cloud‖ in Japan).

1.1 Cloud Computing Definitions

The origin of cloud computing can be traced back as early as the 1960s when John McCarthy proposed the idea of computation being delivered as a public utility^[1] . This idea faded in the

1970s because the technologies were not yet ready. In 2006, Google‘s CEO, Eric Schmidt, publicly uttered the term “cloud computing” and his use of the term triggered a wide discussion within the information technology industry. The term ―cloud,‖ which is a metaphor for both the Internet as a whole and applications that are hosted within it, offers businesses the opportunity for scalable and virtualized control over their organization‘s information system infrastructure^[2] .

Cloud computing has been defined in a variety of ways, each stressing different aspects of the concept. The definition of cloud computing developed by Buyya et al. (2008) highlights the relationship between service providers and cloud service users. They defined cloud computing as a parallel and distributed system consisting of virtualized computers that are dynamically provisioned as unified computing resources based on service-level agreements negotiated between a vendor and its customers. Staten (2008) adopted a relatively narrow definition in which cloud computing is viewed as a pool of scalable and abstracted computing infrastructure for hosting end-user applications capable of being billed on the basis of consumption. From the business model perspective, Armbrust et al. (2009) refers to cloud computing as the applications delivered as services over the internet rather than the hardware and systems software in the data centers.

In this study, I employ the comprehensive definition of cloud computing as used by the National

Institute of Standards and Technology (NIST). NIST defines cloud computing as ―a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.‖ By this definition, cloud computing is not only an evolution of computing technology, but also a new business model for delivering IT-based solutions (Iyer and Henderson 2010). Cloud computing evolves from several precursors such as grid computing (Channabasavaiah et al. 2003; Srinivasan and Treadwell 2005), web services (Berbner et al. 2005), and utility computing (Rappa 2004). As a new business model, cloud computing has changed businesses in several ways—eliminating upfront IT investment costs and the need to maintain in house data centers— but the most significant advantage of cloud computing is cost reduction.

Cloud computing is characterized by a number of capabilities and attributes. Table 1 summarizes some key characteristics from the academic literature and industrial reports of research institutes. These characteristics distinguish cloud computing from other information technologies and show cloud computing to be far more than a simple buzzword.

 

Table 1. Key Characteristics of Cloud Computing
Characteristic		Description		References
On-demand service		Services or computing capacity made available as needed.		Mell and Grance (2011)
	Ubiquitous access		Services can be accessible over the internet.		Armbrust et al. (2009), Mell and Grance (2011)
	High elasticity and scalability		Computing capabilities can be elastically provisioned, and dynamically scaled up or down with demand.		Mell and Grance (2011), Qian et al. (2009)
	Pay-per-use		Users are charged for actual use of service instead of by subscription.		Armbrust et al. (2009), Staten (2008)
	Location independence		Users can access data and services without knowing its physical location.		Iyer and Henderson (2010)
	No upfront commitment		Organizations can start small and increase resources as needed		Armbrust et al. (2010)

 

The NIST delineates three cloud computing service models: Infrastructure as a Service (IaaS),

Platform as a Service (PaaS), and Software as a Service (SaaS). Similarly, Iyer and Herderson (2010) classified cloud computing vendors based on different levels of abstraction. Vendors at the infrastructure level, such as Amazon, provide basic computing resources such as dedicated servers, memory, and storage. Vendors at the platform level, such as Microsoft and IBM, provide a development environment which allows users to create new applications. Vendors at the software level, such as Google and Salesforce.com, offer on-demand software rental. It is worth mentioning that SaaS is different from the ASP (application service provider) model in that SaaS is able to achieve economies of scale by providing one-size-fits-all solutions^[3] .

Cloud computing represents a new fashion in the IT industry. A growing number of companies have labeled themselves as cloud service providers. Among the first of these was the Amazon Elastic Compute Cloud (Amazon EC2). As an early entrant, Amazon is currently the largest provider and market share winner in this market. Table 2 lists some of the major cloud service providers with divided into the three NIST service models. Cloud computing applications have been developed for the need of many areas such as scientific research, healthcare, R&D projects for testing new services, applications and design models, low-priority business applications, and web-based collaboration services (Hand 2007; Staten 2008).

 

Table 2. Cloud Computing Services
Service	Vendors	Description
IaaS	Amazon	Elastic Compute Cloud (EC2) virtualized cloud infrastructure; Simple Storage Service (S3) cloud storage;
	VMware	Vblocks: helps business customers build out clouds based on Vblock packages.
	Rackspace	Rackspace Cloud: Cloud Sites, Cloud Servers, Cloud Files.
	Joyent	Joyent Accelerators (virtual servers)
	3Tera	AppLogic: provides infrastructure solution
PaaS	Google	A platform for developing and hosting web applications in Googlemanaged data centers; supports Python and Java.
	VMware	vCloud: manages applications within private clouds or has them federated on-demand to partner-hosted public clouds.
	Salesforce.com	Force.com Platform (Custom Cloud 2, Development Platform)
	Microsoft	Azure: a Windows-as-a-service platform consisting of the operating system and developer services
	Joyent	OpenSolaris
	IBM	Blue Cloud computing platform: enhances software development and delivery capabilities, particularly in large companies
	NetSuite	SuitCloud: offers on-demand products, development tools, and services
	3Tera	Cloudware: offers applications and storage.
SaaS	Google	SaaS: Web-based communication, collaboration, & security applications including Gmail, Google Calendar, Google Talk, and Google Docs.
	Salesforce.com	CRM (Sales Cloud 2, Service Cloud 2)
	NetSuite	SuitCloud: offers on-demand products, development tools, and services
Notes: Iaas—Infrastructure as a Service; Paas—Platform as a Service; SaaS—Software as a Service.

 

From the perspective of service boundaries, cloud computing can be categorized as public clouds, community clouds, private clouds, or hybrid clouds (Table 3). It has taken some time for traditional IT professionals to both comprehend the benefits of, and establish trust in the concept of cloud computing. Thus, the main users of public cloud computing are small businesses and startups that don‘t have a legacy system to manage or integrate (Staten 2008). It has also proven fruitful to the innovative companies that have taken the initiative to use it. The underlying rationale is that they want to concentrate on core businesses by obtaining IT capacities from public clouds. One successful example is The New York Times, which uses Amazon EC2 and Amazon S3 for its online archives^[4] . However, most big information and communication technology companies build their own private clouds. For instance, IBM, HP, Dell, and CMCC

(China Mobile Communication Corporate) have all built private clouds. Google‘s App Engine and Amazon‘s Elastic Compute Cloud are hybrid clouds because they provide services for both public and private use.

 

Table 3. The Service Boundaries of Cloud Computing
Type	Description	Reference
Public Cloud	The cloud infrastructure is available to the general public and is owned by a cloud service vendor.	The NIST Definition of Cloud Computing (Mell and Grance 2011)
Community Cloud	The cloud infrastructure is shared by several organizations within a specific community that shares a common mission or set of concerns.
Private Cloud	The cloud infrastructure is provided solely for a single organization and is maintained by internal data centers or third parties
Hybrid Cloud	The cloud infrastructure is a combination of two or more cloud models.

 

1.2 The Vision and Reality of Using Cloud Services

Cloud computing has been on the strategic agenda of a growing number of companies because it provides solutions for the specific challenges confronting organizational IT departments. These challenges include the difficulty of capacity planning, the dilemma of balancing time to market against asset utilization, and the business requirement to build IT Infrastructure without detailed business plans and without budget approval (Staten 2008). These issues relate (to some extent) to maintaining in-house data centers or on-premise software applications. As a new IT outsourcing model, cloud computing offers an alternative with two major benefits.

First, using cloud services has significant economic advantages. Cloud computing has the capacity to significantly reduce costs by eliminating both the need for up-front capital investment, and by reducing the necessity to maintain proprietary infrastructure (Jaeger et al. 2008). The pay-as-you-go model converts capital expenses to operating expenses and allows capital to be redirected toward investment in core business (Armbrust et al. 2009). Another reason for its economic advantage is that cloud service providers have extraordinary negotiating power when it comes to hardware pricing, software licensing, and support contracts. For example, Amazon costs $70-$150 per month fully burdened for a server versus an average of $400 per month for an in-house server (Staten 2008). The second benefit is elasticity. Cloud computing can accommodate new services within hours rather than weeks, and can handle workloads closely by dynamically scaling up or down. Armbrust et al. (2009) cited Target and Amazon as examples to show elasticity benefits for both start-ups and well-established large companies: while other retailers have severe performance problems during Black Fridays, Target and Amazon‘s sites were only slower by 50%. This example also suggests another benefit of cloud computing: risk transference, especially the risk of over- or under-provisioning.

Along with the benefits of cloud computing discussed above, there are concerns and uncertainties that hold back current organizations from taking it on board. Jaeger et al. (2008) outlined some of the potential policy issues presented by cloud computing including privacy, security, reliability, access, and regulation. Reliability raises significant problems when service disruptions occur, or data integrity and accuracy is compromised. Security, privacy, and anonymity are the most pressing concerns for cloud users considering the possibility of the unauthorized release of sensitive information, or having content monitored by third parties. Access and usage restrictions may be problems for cloud providers because they have to worry about intellectual property violation and other illegal use by some users. Rochwerger et al. (2009) analyzed factors that mobilize cloud computing as a general-purpose public computing utility, including: ever-advancing hardware and software technology; the continued penetration of internet access; diminishing security and trust concerns on the part of users; and the ubiquity of visualization. In opposition to these drivers, organizations may also face obstacles in the adoption of cloud computing. These obstacles can essentially be categorized into either technical or organizational obstacles (see Table 4). It is argued that some of the technical obstacles may only be valid for a certain time period (e.g. the early stage of cloud computing) but are by no means insurmountable in the future. For instance, the lack of interoperability (Rochwerger et al. 2009) between cloud providers will be solved as cloud computing evolves over time, and standards and protocols for improving interoperability will be developed.

 

Table 4. Obstacles to the Growth of Cloud Computing
	Obstacles	Description	References
Technical Obstacles	Availability of service	Users expect cloud services to be delivered with high availability and reliability	Armbrust et al. (2010); Armbrust et al. (2009), Chow et al. (2009), Dikaiakos et al. (2009), Ranjan et al. (2012), Rochwerger et al. (2009)
	Data lock-in	Users have difficulty extracting data from one vendor to another due to a lack of interoperability
	Data confidentiality and auditability	Sensitive data may be exposed to more security risks and certain auditability requirements need to be met
	Data transfer bottlenecks	Data transfer over the internet could be costly
	Performance unpredictability	Users want quality of service guarantees regardless of workload or use of resources within the cloud
Organiza- tional Obstacles	Differences over the pace of adoption between business and IT executives	While business wants to move forward to a full cloud adoption, IT tends to proceed more slowly because of concerns about security, management, and efficiency	HP Business White Paper (2011), Garrison et al. (2012)
	Different stages of maturity along an adoption continuum	Companies that have adopted virtualization are more ready to adopt cloud computing
	Uncoordinated adoption	Organizations approach the cloud in a fragmented way without an overall adoption strategy
	Inadequate business and technical acumen	The benefits of cloud computing may not be fully realized when cloud computing falls short of expectations because of an inadequate understanding of scope, span and implementation.

 

Industrial surveys were conducted in order to capture the mainstream perceptions and attitudes about the adoption of cloud computing within companies. The Symantec 2010 State of the Data Center Survey reveals that 69 percent of small firms, 85 percent of mid-sized firms, and 79 percent of large firms are involved in cloud computing. BT‘s Enterprise Intelligence survey (2009) reports that over half of the CIOs remain unclear about how cloud services can save money even though cloud services are designed to reduce capital expenditure requirements. The survey conducted by MPS partners in 2009 examines the factors that may influence the use of cloud computing, including more capacity for less money; less upfront investment and lower total cost of ownership; and integration with service-oriented architecture principles. In summary, these surveys suggest that cloud computing today is a nascent but promising paradigm. The major reason why organizations have not jumped onto the adoption bandwagon is that they remain uncertain about its ability to create more efficient IT operations, achieve better use of infrastructure, and save resources.

1.3 Structure of the Dissertation

The dissertation is organized as follows: Chapter two describes the research motivation, questions and scope. Chapter three reviews relevant findings of CSFs of enterprise-wide IT initiatives and IT-enabled organizational change from previous studies. A new research framework will then be proposed that conceptualizes the bi-directional relationship between IT departments and cloud computing implementation. Chapter four delineates the research methodology; and the overall paradigm and research method—including data collection and analysis—of two studies will be discussed. Chapter five presents an essay of investigating factors that influence the success of cloud implementation within organizations. Chapter six presents another essay that will examine the impact of cloud implementation on the IT function. In Chapter seven, key findings will be summarized, followed by a discussion of the theoretical and practical implications of the research, concluding with a few thoughts on both the limitations and possibilities for future research.

[1] http://computinginthecloud.wordpress.com/2008/09/25/utility-cloud-computingflashback-to-1961-prof-johnmccarthy/

[2] MPS 2009 survey on ―The Future of Cloud Computing‖

[3] http://www.cio.com/article/109704/Software_as_a_Service_SaaS_Definition_and_Solutions

[4] http://open.blogs.nytimes.com/2007/11/01/self-service-prorated-super-computing-fun/

CLOUD IMPLEMENTATION IN ORGANIZATIONS: CRITICAL SUCCESS FACTORS, CHALLENGES, AND IMPACTS ON THE IT FUNCTION

Sharing is caring!