The only positive critical influence of method was the use of prototyping to build a large system in phase I on time and under budget (C.2.1a). However prototyping and the requirements study also had important positive influences on the development (C.1.3a, C.3.1c) and on the careers of both users and developers (C.2.5d).
Despite its positive influences, prototyping also had both critical and important negative influences on both construction (C.2.1a, C.2.2a) and delivery (C.3.1k) and the requirements study also had a critical negative influence by specifying a first phase that was too large (C.1.4b). Project management, installation and data conversion also had negative impacts on development (C.2.4a, C.3.1h, C.3.2e). These negative influences were not just caused by tasks being undertaken badly or inappropriately but also because social influences overwhelmed method.
During development some appropriate methods that were used were either ineffective or ignored resulting in critical or important impacts on development. Project management and the IT Strategy both come in this category (C.2.4a, C.1.2a). Quality Assurance was also largely ineffective or ignored but the impact was rarely important or critical (C.1.2e, C.2.3e).
There was a view that the IT Strategy was only undertaken to justify a system which Tim had already decided to implement (C.1.1) and Colin had already designed the architecture (C.1.3b). Maybe it was this or a more general industry attitude to IT strategies (S.8.8) but the results of the IT Strategy were largely ignored (eg replacement of Q&A system, difficulty of getting underwriters to use the system). It would have been better if the IT Strategy had been heeded or failing this maybe time and money could have been saved by doing a reduced version to satisfy X-Group.
The quality assurance functions in GIS were also largely ignored by the developers (C.2.3e, C.2.5e). This was partly because they were out of step with the prototyping approach adopted by the team. Attention to the Post Implementation Review for phase I might have resulted in the leadership issues of phase II being addressed quicker but the worst consequence of the lack of impact of the quality assurance function is that it wasted time and money.
The use of prototyping had a large impact on the development of the Multinational system, both positively and negatively. A better understanding by the team that method can impact on functionality (C.2.1a), design (C.2.2a), testing and thus reliability (C.2.3c) and scheduling (C.2.4a) would have led them to view prototyping more critically. This might have enabled them to gain the benefits from prototyping without suffering so many of the drawbacks that they had to abandon it in favour of more structured approaches.
© Clare Tagg 2000