IJSIM 7,1 74 Factors affecting information systems��� success Grafton Whyte and Andy Bytheway Cranfield School of Management, Cranfield, UK Introduction Success and failure with information systems In the history of information systems in business it is probably true that there has been more failure and disappointment than success. Work by Butler Cox Foundation (1986), Galloway and Whyte (1989) and Lyytinen (1988) suggests that one in two information systems development projects will not lead to successful systems. If end users[1] are asked what they think about systems that have been delivered to them they are likely to display a degree of indifference. Experts working with information technology often perceive successful systems to be concerned with the successful use of the latest technology and not with the relevance of that technology to the users��� needs. Project managers see success as a project which has been completed in the face of inevitable difficulties, which may in fact have been self-inflicted. Users seem to have no clear means whereby they can judge the success of a system, although they are more easily able to determine that the desired business benefits are not being delivered by the system. These are signs of an immature discipline which is probably not yet understood and which therefore can not be managed properly. IT management and systems project leaders need to understand their users��� view of success and the factors that affect its achievement. The technological and project management viewpoints need to be balanced by other considerations which take appropriate account of the users��� needs and expectations. Purpose of the paper The purpose of this paper is to report new work which has investigated ideas in the service management discipline, and tested their applicability to the analysis of information systems��� success. The paper discusses the concept of success and failure in the provision of information systems and reviews previous work that has addressed these issues. It then introduces the new work, which looks at the problem from a service perspective and uses repertory grid techniques and statistical analysis to develop a new set of service attributes with which to understand information systems��� success. The use of the attributes is illustrated using two examples. Previous work Previous work adopts different perspectives on the problem of successful information systems delivery, leading to quite different views as to its possible International Journal of Service Industry Management, Vol. 7 No. 1, 1996, pp. 74-93. �� MCB University Press, 0956-4233 Received January 1994 Revised October 1995

Factors affecting information systems��� success 75 improvement. If we are to achieve a proper balance, all relevant perspectives on the problem need to be considered. Product, process and service perspectives Three perspectives (see Figure 1) which are evident in the literature and which may be important are: the product which is delivered to the users (for example the software and hardware systems, user documentation and training courses) the process that creates the system (traditionally including systems analysis, technical design, program coding, testing and final handover) and the service which deals with the softer issues (answering questions, dealing with problems, and generally addressing the concerns and aspirations of users). If asked, most IS (information systems) managers would probably already consider themselves to be delivering a service. However, the evidence is that these managers spend most of their time monitoring aspects of their operation which are more concerned with the product and the process, and which have little to do with service. Systems providers and software houses tend to fall into the same trap. It follows that service management is a discipline that may be of direct help, and which is worthy of examination in the context of difficulties with information systems. Previous information systems research Much early information systems research has chosen to deal with the product viewpoint. Studies into the determinants of success and failure have focused on the more observable, tangible attributes and characteristics of systems products, such as system response times, data volumes and the extent of systems usage. Increased systems complexity, the increasing number of unsuccessful systems and a growing systems development backlog led to a Figure 1. Perspectives on information systems Service perspective Process perspective Product perspective Information system

IJSIM 7,1 76 shift in attention from the product to the process viewpoint. Studies again concentrated on the more tangible attributes such as the number of errors occurring within the process, the level of user involvement and the milestones at which user approval is given. A review of the information systems literature suggests the following as being the main causes of difficulty: over-optimistic estimates that subsequently lead to the system being delivered late and over budget (Brooks, 1975 Galloway and Whyte, 1989 Keen, 1987 Rademacher, 1989) ill-defined project objectives, mostly arising from uncertainty regarding the business needs to be satisfied (Keen, 1987 Lyytinen, 1988 Rademacher, 1989) poor communications between users and the development staff (Illes, 1990 Whyte, 1987) and lack of user commitment to the project and system (Keen, 1987 Tait and Vessy, 1988) the technical limitations of a system and systems which are unfriendly and inflexible (Galloway and Whyte, 1989 Lyytinen, 1988) and the use of inexperienced staff to develop systems (Illes, 1990 Rademacher, 1989 Keen, 1987). These causes of difficulty are largely within the realm of project management, and we have to ask whether the list is actually complete. Work by Galloway and Whyte (1989) and Jones and Kydd (1988) suggests that these causes may actually be symptoms of underlying problems relating to uncertainty (the lack of information), equivocality (the absence of clarity, or excessive ambiguity in the project) and internal inconsistency (between the key elements of an information system project ��� for example between the users, developers, tools and technology used in it). Miller (1989) also casts some doubt on the ability of product and process approaches to measure information system success. He suggests that process outcomes may not be able to be successfully measured. On the product side he (along with a number of other authors, for example Melone, 1990 and Srinivasan, 1985) looks at systems usage as a key indicator, but concludes that it is beset by a number of complicating issues such as: whether use of the system is mandatory or discretionary the influence of viable information alternatives the effect that user experience has on usage and the extent to which obtained information is actually used. It is also possible to classify research according to the analytical approach taken. For example, Kydd (1989) argues that there are four broad but significantly overlapping categories of analysis, and these are supported and illustrated by many other authors: technical approaches (further illustrated by Brooks, 1975 Casher, 1984 Hughes 1986 Huling, 1987 Kaniper, 1986 Kuzman, 1989 Morreale, 1985 Simon and Davenport, 1987 Viskovich, 1988 Yaffe, 1988) behavioural approaches (further illustrated by Carroll, 1982 Doll, 1985 Jones and Kydd, 1988 Necco, 1989 Pinto and Slevin, 1987 Tait and Vessey, 1988) organizational approaches (further illustrated by Ginzberg, 1980 Rademacher, 1989) and interactive approaches (further illustrated by Kydd, 1989 Skyrme and Earl, 1990).