A critical review of cloud computing: researching desires and realities

Abstract

Cloud computing has become central to current discussions about corporate information technology. To assess the impact that cloud may have on enterprises, it is important to evaluate the claims made in the existing literature and critically review these claims against empirical evidence from the field. To this end, this paper provides a framework within which to locate existing and future research on cloud computing. This framework is structured around a series of technological and service ‘desires’, that is, characteristics of cloud that are important for cloud users. The existing literature on cloud computing is located within this framework and is supplemented with empirical evidence from interviews with cloud providers and cloud users that were undertaken between 2010 and 2012. The paper identifies a range of research questions that arise from the analysis.

Appendix

Research methods

We undertook over 50 semi-structured interviews with senior staff within a range of cloud-based organisations from across the supply chain between June 2010 and May 2012. We interviewed providers of cloud infrastructures and services, system integrators and users of cloud services. In terms of roles, we spoke to CEOs, CIOs, marketing managers and service directors. Interviews were normally undertaken by one person and were typically held over the telephone. They typically lasted an hour, with some running to over 2 h. Over 30 of these interviews were recorded and transcribed. A detailed analysis of the transcribed interviews was undertaken and the insights gained from the coding process were checked and challenged with the available literature and notes taken during other, non-transcribed, interviews.

The transcribed interviews were then coded by two of the authors. An inductive approach was adopted and no particular theory was identified a priori to frame and guide the data analysis although we drew heavily on our knowledge of IT outsourcing. In contrast through the analysis process a number of relevant concepts emerged which aided in making sense of the data. After much iteration, these ended up forming the desires framework presented above. Initially codes were used to simply classify each element (‘quotations’) of the interview. As the interviews were being coded, a parallel process of consolidation took place.

The first step towards consolidating codes into analytically distinct segments that can be examined together both within and between interviews involved tidying up the initial codes, for example by combining codes that covered the same concept but were labelled slightly differently. This process of analysis was based on and contrasted with themes from the cloud and outsourcing literatures which were being reviewed concurrently (Eisenhardt, 1989).

The process involved an iterative reading, coding and cycling through the codes. The validity of the coding and analysis was constantly checked by searching for counter examples and nuances in the text and codes. Through this process a series of higher-level thematic codes emerged.

Finally, a selection of the coded quotations was selected for use in the paper (Golden-Biddle and Locke, 1993). The selection process was guided by the need for a coherent narrative flow in the paper, so quotations that best illustrated the point being made were used. Other coded paragraphs would frequently feed into the underlying argument, even if they were not being directly quoted.

Limitations of this research

Despite being grounded in both the relevant literature and current empirical practice, the research has some limitations. As an attempt to categorise the complexity of existing, emergent practice, where different stakeholders may emphasise different aspects of cloud for their own purposes and where some have rebadged existing provision in terms of cloud, there will inevitably be some areas that are less well represented in the framework than others. We reflect on three such areas of under-representation below.

Potentially significant in this regard is the absence of a formal location for ‘privacy issues’. We recognise that privacy and data protection issues are frequently cited as a reason for failing to move to the cloud and while there are a range of technological measures that can be resolved to address many of the existing privacy concerns, we subsume discussion of these within the equivalence desires of the framework. That is, what level of performance equivalence is required to satisfy data protection concerns and what consequences does this particular desire have on the particular form of cloud adoption (e.g. hybrid models).

Another important area of research relates to the skills necessary to manage cloud adoption within an organisation. For example, what kinds of architectural skill sets are required to link internal processes with the cloud, what sort of contract management skills are necessary to maximise the efficiency of cloud contracts?

Finally, cloud opens up challenging governance issues in terms of oversight of the cloud providers that an enterprise interacts with to deliver its business services.