تطبیق دو روش برای عوامل مهم موفقیت : مورد خدمات مشترک در بخش دولتی

Shared services have been embraced by the private, and increasingly, the public sectors. Yet implementation has often proved to be difficult and the factors which are critical to success are not yet well understood. In this paper existing research in the area of critical success factors (CSFs) is examined and it is suggested that that research actually covers two distinct phenomena. One approach is focused on identifying the factors required for a specific individual to achieve their outcomes. The second on determining the general success factors for implementing a project of a certain type. A reconciliation of the two approaches is proposed within a framework that distinguishes between three different types of CSF – outcome, implementation process and operating environment characteristic. A case study of a project to implement shared services in the Australian public sector is examined using the reconciled framework. The case shows that the reconciled approach by including, and differentiating between, outcome, process and operating environment characteristic factors provides a richer and more complete picture of requirements. Further benefits from the synthesis are also highlighted including that factors are a mix of universal and context specific, different perspectives on factors exist and not all environmental factors perceived to be critical have to necessarily be present.

While many definitions of shared services with slightly different nuances exist – see Schulz and Brenner (2010) for an extensive review – the fundamental essence remains broadly the same. A shared service is one where the provision of a back office service – such as payroll processing, accounts payable or foundational IT services – is consolidated within a single area of an organization (Longwood and Harris, 2007, Quinn et al., 2000 and Ulbrich, 2006). It typically replaces arrangements where there is a duplication of efforts among different business units. Initially the focus for the introduction of shared services was the private sector and there are some well-known success stories. Bergeron (2003) provides the example of Bristol Myers Squib's global business service unit realizing annual savings of $1.5 billion through shared services while Cecil (2000) cites Ford being able to reduce its finance department staffing from 14,000 to 3000. More recently the potential of shared services in the public sector has started to be investigated (Janssen et al., 2007, Niehaves and Krause, 2010 and Ulbrich, 2008). Janssen and Joha (2006), for example, suggested that shared services can offer multiple benefits such as reducing costs, improving access to innovation and allowing an increased focus on core operations. Yet, it is also being realized that shared services success is not guaranteed. Wagenaar (2006), for example, detailed the failure of a major shared services initiative in the Netherlands. More recently the government in Western Australia cancelled a project, based on the implementation of a common ERP, to merge finance and payroll processing services across multiple departments. The project, started in 2007, had been estimated to cost AU$82 million and deliver annual savings of AU$57 million. However by the time of termination AU$401 million had been spent (Kerr, 2011). Shared services are of specific interest from an information systems (IS) perspective not only because of their direct potential with regard to IT services but also because – as highlighted above – many other services such as payroll processing or accounts payable are dependent upon IS for their delivery. The IS academic discipline increasingly recognizes the merit of looking beyond the IS itself to examine the functions and activities that are enabled by IS – see for example Willcocks, Lacity, and Cullen (2007), Leonardi and Bailey (2008), or Hagel and Seely Brown (2001). Given the potential benefits and risks associated with implementing shared services this paper seeks to provide a framework that can enable a proposed project to be rigorously assessed before the decision to proceed is taken. It has been suggested previously that the implementation of shared services represents a strategic decision ( Bergeron, 2003 and Su et al., 2009). One of the criteria suggested by Eisenhardt and Zbaracki (1992) to classify decisions as strategic is the scale of the resources being committed and there appears little doubt that the shared service examples provided earlier satisfy that criteria. As such the strategic decision making literature could be a good starting point to look for a suitable framework. Strategic decisions typically are not simple with clear objectives, boundaries and considerations but are rather a complex “mess” (Ackoff, 1979). Often multiple criteria and variables need to be incorporated and balanced (Belton & Stewart, 2002). Yet it has long been recognized that decision makers are poor at handling such complexity ( Duhame and Schwenk, 1985 and Malhotra, 1982). Indeed there is a long tradition of research arguing the need for simplification. Simon (1957) for example argued that “the description of the real-world situation is radically simplified until reduced to a degree of complication that the decision maker can handle” (p. 170). Miller (1956) suggested that decision makers have a limited information channel capacity and can only effectively incorporate consideration of a limited range of variables. Research by Hodge and Reid (1971) argued the need for focus finding that the provision of irrelevant information to a decision impeded the ability to identify and assess relevant information and thus reduced decision making performance. A focus on critical success factors (CSFs) could be an appropriate simplification vehicle. While Daniel (1961) is credited with first developing the concept of CSFs, Rockart (1979) popularized it. He defined CSFs as the few key areas in which favourable results will ensure successful competitive performance. The intention was to identify the structural variables that most contribute to the attainment of strategic goals and objectives. The CSF approach was originally developed to assist managers determine their information needs – but has since been ported to a wide variety of contexts. It has, for example, been used to examine business process management (Trkman, 2010), the adoption of e-banking (see Shah & Siddiqi, 2006 for an extensive review), e-Government implementations (Sagheb-Tehrani, 2010), the deployment of mobile applications (Al-Hadidi & Rezgui, 2009) and the alignment of information systems and business plans (Teo Thompson & Ang James, 1999). It does not appear however to have been applied to shared services. It is suggested here that taking a CSF focus could be beneficial when making a decision regarding whether or not to introduce shared services. However before applying CSFs it is seen as necessary to provide some clarity regarding the approach itself. The work of Rockart (1979) on CSFs was further developed by Bullen and Rockart (1981) but appears to differ from much of the subsequent work in the area in three key ways. Firstly, Bullen and Rockart (1981) concentrated on uncovering what outcomes were necessary to achieve desired objectives. Specifically they were interested in identifying what information needed to be delivered to executives in order for them to do their job. Later work however has typically focused on determining the requirements to implement projects. Secondly, while Bullen and Rockart (1981) suggested that CSFs are context specific – “CSFs are the particular areas of major importance to a particular manager in a particular division at a particular point in time” (p. 3) – other research often appears to be oriented, either implicitly or explicitly, toward identifying universal factors. Thirdly, and relatedly, Bullen and Rockart (1981) examined CSFs from the perspective of the individual seeking to identify those of specific executives. Later work by contrast has focused on CSFs from a broader, anonymous, perspective identifying general sets associated with projects, such as successful alliances (Rai & Borah, 1996) or systems implementations (Wixom & Watson, 2001), rather than individuals. In short, there are two approaches. One focused on outcomes individuals and specificity the other on process, projects and universality (see Table 1). Interestingly the second approach does not appear to have emerged as a result of criticisms of the original but simply as a different interpretation of the underlying concept of critical success factors. Even Rockart appears to have participated in this reinterpretation to some degree (see Rockart, 1982).Rather than choosing one approach or the other this paper proposes that the usefulness of CSFs would be enhanced if both perspectives were to be combined. Such a reconciliation will allow for consideration of the necessary outcomes of an initiative and the requirements for delivering it, individual and overarching perspectives and a recognition that some factors may tend toward being universal in nature while others will be more context specific. The paper also puts forward two enhancements. Firstly, that a third type of CSF – relating to an organization's operating environment – that has been implicitly considered by both approaches should be explicitly separated out. Secondly, that including missing CSFs in deliberations would be beneficial. The paper contributes to the literature in two ways. Firstly it identifies specific factors perceived as being critical to a successful implementation of shared services which may provide guidance to others. Secondly, and perhaps more important in the long run, is the reconciliation of the two different CSF approaches. The remainder of the paper is structured as follows. The next section presents the rationale for the proposed CSF reconciliation. Section three briefly describes a case study of a project to implement shared services demonstrating the application of the reconciled CSF approach and identifying specific CSFs. Section four discusses that empirical work and the final section provides conclusions, limitations and suggestions for future work.

Historically two distinct approaches to using CSFs to support decision making in complex messy situations have developed – one focused on outcomes, individuals and specificity the other on process, projects and generalizability. The second approach does not appear to have emerged as a result of criticisms or refinement of the first but rather from a different interpretation of the underlying concept. In this paper a framework reconciling both approaches by classifying CSFs by outcome, implementation process or operating environment was proposed. 5.1. Implications for research and practice From a CSF perspective, this paper illustrates that it is possible, and useful, to reconcile the two CSF approaches and identify and classify CSFs according to the three domains of outcome, process and operating environment. Furthermore, and beyond this primary aim, the paper enhances understanding of the underlying CSF concept in three ways. Firstly, the case shows that consideration of CSFs should not be restricted to factors that are present but should also include factors that are absent. Not all projects take place in perfect circumstances. This does not mean that the shortcomings of those circumstances are not recognized. This is perhaps particularly relevant to researchers seeking to identify factors in retrospect. They need to ensure they do not just ask what factors were critical to success but also what factors, if they had been present, would have enhanced success or made it easier to attain. Secondly it reinforces that individual CSFs can differ with regard to whether they are universal or context specific. Thus it is necessary to take a middle ground that neither sees previously identified CSFs as definitive – as authors such as Nah et al. (2001) would suggest – or largely irrelevant – the view of Bullen and Rockart (1981). Instead such CSFs serve as a foundation or template that may require adaptation as it is applied in new contexts. Thirdly the research highlights the possibility that the meaning of any given CSF may be open to interpretation and that therefore care needs to be taken when using seemingly common terminology. This may be especially important when porting a set of CSFs to other initiatives or organizations. From a decision making perspective the ability, and value, of the CSF approach with regard to the focusing attention on a limited number of areas is further demonstrated. Even including project and individual factors only a relatively small number of critical factors were identified. By giving more explicit attention to the existence of absent CSFs the new approach also better reflects the messy reality of decision making – where options are rarely perfect and the need to compromise common. From the practical perspective of shared services in the public sector, the contribution of the research is threefold. The identification of actual critical success factors will help other agencies and entities better understand the types of things that might be relevant in their own context as they consider moving to implement shared services. Secondly incorporating a three dimensional view of CSFs – outcome, process and operating environment – should lead to a more comprehensive process of surfacing and assessing CSFs and contribute to agencies making more considered choices regarding whether to implement shared services or not – regardless of the actual CSFs and their relevance or transferability to other contexts. Thirdly, where multiple absent factors are identified this can serve as an early warning signal with regard to whether an initiative is appropriate. Further CSF research should take into account the distinction between outcome, process and operating environment. 5.2. Limitations There are several limitations to the research. Firstly, the views of the interviewees are forward looking perceptions of factors that they believe will be critical to the success of the shared services project. All the interviewees are multi-year veterans both in their profession and Education which provides some grounding for those perceptions. However the project is in its nascent stages and it is therefore not possible to determine if the factors identified will actually contribute to success. Second – and this is a limitation in common with much the CSF research (see for example Bergeron & Bégin, 1989) – the factors are subjective and lack any attempt at quantification. Subjectivity means that there is no clear way of determining the precise relationship between a factor and success. Indeed, as Leidecker and Bruno (1984) accept, a factor might be identified based on nothing more than biased opinion (though they argue strongly in favour of the value of the approach). Regarding quantification how does one know or determine to what degree processes should be standardized? Or what represents an evolutionary approach to roll out? Again this problem has long been acknowledged but is not seen as making the use of CSFs valueless (Leidecker and Bruno, 1984). Third interviews and the resultant factors were identified by a single sub-group of the organization's employee population – senior management. It is possible, perhaps probable, that interviews focused on a different sub-group – for example teachers or administrative staff would yield different factors and even objectives. However as Walsham (1995) suggested researchers do not need to strive for a single or universal interpretation. A complete picture of a phenomenon is unlikely to be constructed through a single piece of research, it is likely to be built over time by combining multiple perspectives each of which adds different, but equally interesting, insights (Brunsson, 1982). Given the centrality of senior management to significant organizational change of the type the introduction of shared services represents (Tushman & Romanelli, 1985) their perspective seemed a sensible starting point. Fourth, and finally, the research is based on a single case. There is no guarantee that the findings will be generalizable. This relates both to the usefulness of the framework itself and to any shared service specific insights. Again though it should be recognized that this research is intended to be a starting point and not an end. Future work, refining and extending what has been presented here, will ultimately reveal its usefulness. 5.3. Future research Future research could be conducted in a four main areas. Firstly to conduct additional case studies of shared services to gain a better understanding both of the extent to which the framework is relevant to other organizations and the degree to which the specific factors identified are generalizable or context specific. Assuming such work yields positive results it would then be useful to conduct quantitative research that could test hypotheses concerning relationships between factors and success. A second avenue of research, building on the first, would be to compare the factors identified empirically in this research in more detail with those that appear in the literature for other domains. It would be interesting to reflect on the similarities and differences. In a different vein the research here suggests it would also be useful to improve understanding of the CSF concept itself. The research does not shed much light on whether all CSFs are of equal importance. Improved understanding here is critical though given that some may be absent. Are there some factors, for example, that a project absolutely cannot succeed without. Furthermore, given this study was focused on the beginnings of the project, it raises the question as to whether CSFs are constant over its duration or change – and if they change over time why this is the case. Two types of change are possible. Firstly, what is expected to be important could differ from what actually is. Secondly, what is important for a project's planning differs from what is for its execution.