Driving IT projects to success: stakeholders’ importance: an artificial neural network model to demonstrate the potential of using stakeholder characteristics in IT projects’ success estimation

Abstract

Technology is all around, progressively present with each passing day. Companies recognize the usefulness of technology in business, leading to a growing number of Information Technology (IT) projects development. Due to its increasing scope, IT projects are getting more and more complex and expectations on their results are at an all-time high. At this rate, there is no telling where this complexity will lead, nor if expectations can be met. The development of IT project, or projects of any kind, is always met with unforeseen risks. Therefore, models that aim to estimate the success of these projects have been emerging. Some of these tools have fallen upon the bias of only taking into consideration a few project management variables for forecasting success. This may lead to inaccurate estimations, from the point-of-view of the several stakeholders. Considering the intricacy of IT projects, and the several aspects that influence them, advanced statistical models are required to give rich insight into projects’ outcome. On the other hand, project success cannot be fully determined if the stakeholders’ points-of-view are not taken into account. In other words, the success index of a project must be estimated having stakeholders taken into consideration. In order to support the mentioned concerns, a predictive model using Artificial Neural Networks was developed. Projects and stakeholders characteristics are defined, along with projects’ success criteria as inputs of the model, generating success indexes by budget, time and scope performance, as well as an overall success index as outputs. This dissertation adds to the current literature on the subject, by demonstrating the importance of stakeholder characteristics in project estimation and paving a pathway for the further exploration of the model developed. Thus making a first step into building a prediction tool to help mitigate the current risks of IT projects and software development.