Predicting the future and learning from the past: Developing a framework for evaluating success in IT projects

Research output: Book/anthology/dissertation/reportPh.D. thesis


    Anne Færge Abell
Investments in information technology are significant and seem to follow an ever-increasing trend in modern companies in general and in the LEGO Group specifically. Projects take up a substantial amount of the human and financial resources spent on IT – and for good reason: the project portfolio is one of the most important elements of strategy implementation in any organization. From this follows that IT projects should be immensely interesting from top management perspective since projects are extremely resource consuming and significantly affect the organization’s chances of survival. Failing in managing IT projects is potentially catastrophic while succeeding with IT projects can potentially ensure organizational prosperity. Based on this, it seems obvious that any organization would monitor IT project success very closely. However this is neither the case in the LEGO Group nor in general. Instead, organizations tend to take a very unstructured approach to evaluating IT projects. This in turn render impossible comparison of projects and aggregation of information on project success, and as a consequence it is very difficult to learn from the past and predict the future of new projects.

This PhD dissertation remedies this issue by conceptually defining the concepts of ‘success’ and ‘IT project’ and normatively create a framework by operationalizing the definitions allowing for cross-project comparison of success of an IT project to a reference class, which enables both learning from the past and prediction of the future. Two frameworks are defined generically and then tested in practice in the context of the LEGO Group; the Extended Diamond Approach (EDA) for characterizing projects and dividing them into reference classes, and the 4x5 Framework for evaluating success in a standardized manner which makes cross-project comparison possible. The dissertation contributes on a theoretical level by conceptualizing and operationalizing the two key concepts, on an empirical level by adding to the academic discussion of IT project success rates by providing insight from the LEGO Group, and finally practically by illustrating how the LEGO Group can utilize the frameworks in the processes of predicting and learning. Evaluation of success in IT projects is with the EDA and the 4x5 still difficult, un-tangible, costly, complicated, and done with reluctance, but the dissertation gives evidence to why and how the outcome of the evaluation makes it worth it in the end.

It is found in the study that the sample of IT projects studied is quite characteristic for an internal IT department functioning as support to the primary value chain in a production company: the projects are on average of low uncertainty but high urgency. The study showed that the organization takes a structured approach to evaluating success and is relatively successful compared to the results reported in academia in general, but the study also revealed areas for improvement. The most fundamental is, perhaps, that the link between the organization’s overall strategy and the projects is weak. The two frameworks were additionally used in combination to illustrate how it is possible to predict the likely success of new projects based on the sample as well as test hypotheses about the organization’s project management practice.
Original languageEnglish
Place of publicationAarhus
PublisherDepartment of Business Communication, Aarhus School of Business and Social Sciences
Number of pages292
StatePublished - 28 Apr 2016
SeriesPhD Dissertation

See relations at Aarhus University Citationformats

Download statistics

No data available

ID: 99628954