Combining Business and Network Simulation Models for IT Investment Evaluation

Although the inherent interrelationships between business processes and the underlying Information Technology (IT) infrastructure imply that the design of these two organisational facets should be performed in parallel, this doesn’t appear to be the case in practice. For example, simulation is being extensively used in both the business and the IT domains, albeit in a disjoint fashion. In this paper, we investigate the potential of integrating different simulation models to facilitate concurrent engineering of business processes and Information Technology. We present an example case of such integration and propose a holistic approach to IT investment evaluation by simulation. Drawing on the case findings, we identify a number of pertinent issues and articulate future research directions towards the integration of simulation usage in the business domain. 1. The Problem of IT Investment Evaluation Many organisations undertake business improvement efforts (Business Process Reengineering, Continuous Process Improvement, Total Quality Management, and others) in order to reduce costs, streamline operations, and gain strategic and competitive advantages in the marketplace. Information Technology has been acknowledged as a major enabler of business change [14, 36] as it allows organisations to adopt operating practices that would not have been possible without the opportunities offered by new technologies [10, 22]. However, organisations may face various problems when assessing new business processes to adopt and the IT applications that will support these processes. IT can affect business operations in diverse ways and the ultimate impact on business performance depends on various intertwined factors that can render IT Investment Evaluation a difficult task [3, 13, 34, 37]. Many of these factors relate to what might be termed as ‘the problem of measurement’. Many of IT benefits refer to strategic and competitive advantages that are inherently difficult to quantify, are realised in the long run and may be 0-7695-0001-3/99 $10 associated with a high degree of uncertainty and unpredictability, and are indirect to businesses and therefore indistinguishable from many other confounding factors (for example people, processes, and strategy). In this paper we will be concerned with the measurement problem in the context of ex ante IT evaluation, i.e. the evaluation of a proposed system before its implementation, as opposed to the ex post appraisal of an existing system’s performance. In this context, IT evaluation will be treated in a manner somewhat different from the ‘typical’ evaluation practices in the context of structured IS development approaches (for example, the system development life cycle) where evaluation is typically concerned with measuring the degree of technical success of an IT project after its finish, rather than assessing the organisational fit of a proposed system before its implementation. Various approaches and techniques have been proposed to assist organisations in evaluating IT investments (examples of such methods can be found in [4, 9, 20, 23, 26, 34, 35, 37, 40]). Despite the availability of these techniques, Ballantine et al [2] has found that most companies use simple accounting techniques, notably cost-benefit analysis (CBA) or Return on Investment (ROI), in order to decide whether to proceed with a certain IT investment or not. However, the application of any financial technique for assessing a particular IT investment is hindered by a difficulty in identifying and measuring the expected benefits of the proposed investment [7, 13]. Indeed, the inability to express in monetary terms the intangible, indirect, or strategic impact of IT on business performance may be one of the reasons behind the high failure rate of many BP change programs or large IT projects [1, 21, 22]. 2. Simulation for IT Investment Evaluation According to Davenport and Short [11], although business process design and Information Technology are natural partners, their relationships have never been fully exploited in practice. One the one hand, it is naturally expected that the choice of a particular way of conducting