مقاله سال ۲۰۱۷

Organizations undertake projects as vital means to implement strategy and realize value (Chih and Zwikael, 2015). However, a great proportion of projects fail, e.g. in 2015 only 29% of software projects are successful, with 52% of the projects canceled and 19% failed to deliver the expected results (Dannis, 2015). One of the main reasons for this situation lies in today’s rapidly changing environment. Uncertainties, which cannot be fully estimated and often involve ‘unknownunknown’ events like evolving strategy, introduction of new technology and resource conflicts, have impact on project implementation and force the deviation of perceived value from expected goals. Thus even if organizations make great efforts to maintain accurate evaluation of the uncertainties and devise well-designed project plans, project plans never perform in the predicted way, and if the deviation grows, projects will fail. Under these circumstances, effective project implementation processes that consider dynamism under uncertainty should be explored.

The conventional project implementation methodologies follow a linear logic to bring projects ‘back on track’ with respect to the pre-determined operational plans (Hazır, 2014), whereas recent research suggest that the on-going project is an open system, with both its goals and implementation status evolving (Lee et al., 2006; Aritua et al., 2009). In the dynamic environment, projects have to continuously interact with their implementation context, adapting and evolving requirements throughout the system’s lifetime to cope with uncertainties (Locatelli et al., 2014). Thus project implementation process should involve not only foresight, but also remedial actions in response to unexpected changes, requiring the combination of both proactive and reactive activities. Some research refer to this perspective as ‘bounded planning’ and ‘interactive problem solving’, and claim that the value of a project is not well-known in advance, but being defined and updated with uncertainty prevailing (Engwall, 2003; Ahern et al., 2014). Moreover, the non-linear interdependencies between different project components make the system more complex. These interdependencies may form multiple feedback mechanisms, with which even small variation in individual components may diffuse into serious crisis on the overall project (Williams et al., 2003). Thus without looking into the comprehensive system structure, the effects of both uncertainties and remedial actions on project outcomes are difficult to understand.

Since human activities dominate the project implementation processes, including perceiving and reporting the changes, evaluating the remedial action proposals and making reactive decisions, we should look beyond the ‘hard’ operational data and focus more on ‘soft’ factors like stakeholders’ perceptions and behavioral biases (e.g. reporting errors and escalation of commitment) (Meyer, 2014). System dynamics (SD) modeling is applicable here, which can combine both ‘soft’ and ‘hard’ paradigms in the following way (Pidd, 2009; Rodrigues, 2000): Firstly, when formulating the SD model, multiple stakeholders have to coordinate on the central structure of the system (main components, links and feedback loops) and then draw up the causal diagram. This procedure promotes the organizational learning and provides insights into project implementation. The second procedure is the computer-based simulation, which provides explicit suggestions such as what the possible remedial actions would bring, and when and how to intervene. At this procedure, the SD model can use operational data monitored by conventional methods (Lee et al., 2006).