Abstract
It is commonly stated that information systems continue to be plagued by persistently high rates of failure. However, we argue in this paper that the relationship between success and failure is more complex than usually assumed, and based in the different expectations that different stakeholders have of a development effort. The expectation failure concept of Lyytinen and Hirschheim is used as a starting point for discussion leading to the introduction of a new concept that we call ‘defining characteristics’. We then proceed with a discussion of the implications of this new concept for ideas about success and failure and use a case study conducted by the second author to illustrate these ideas.
Table of Contents
Information systems success and failure have been much discussed in the literature for many years (e.g. Brooks, 1974; Davis et al, 1992; DeLone and McLean, 1992; Fortune and Peters, 2005; Lucas, 1975; Lyytinen and Hirschheim, 1987; McFarlan, 1981; Sauer, 1993). It seems, however, that the general assumption in all of these cases has been that the two concepts, success and failure, are inverses of each other. That is, a failure is by definition not a success and vice versa. While this appears natural enough and in accord with common sense, we argue against this view in this paper contending instead that the criteria that distinguish between success and failure are best regarded as being independent of each other. If this is admitted then it suddenly becomes possible for an information system or indeed any other project to be not only a success or a failure in the usually understood sense, but also both at once, or neither.
Contrary to common practice in much of the information systems literature dealing with this kind of topic, in what follows we take ‘success’ and ‘failure’ to be ground terms. We make no direct attempt to define them, for reasons that will become evident below. Moreover, information systems researchers are increasingly recognising the varying needs and expectations of different stakeholders in information systems development efforts (e.g. Seddon, 1997, 1999; Rai et al, 2002; Fortune and Peters, 2005) and, as we shall emphasise below, this has corresponding implications for their potentially disparate views concerning the eventual success or failure of these efforts.
We further argue that what is critical is to identify, for different stakeholders, the specific factors that will distinguish for them between success and failure. This is consistent with Seddon et al (1999) who say that in ‘a world of conflicting human interests and vastly different systems, different sharply-focused measures of IS effectiveness are likely to be needed for different purposes’. Effective elicitation of these factors in the early stages of system development would, we contend, significantly assist developers by giving them an even better understanding than more traditional techniques provide, of what they should achieve, and also avoid doing, in their subsequent efforts.