No such thing as a perfect system

Michael Bronzite's text is oracular in style and occasionally the rhetoric of persuasion overtakes his objectivity. The title promises a strategic framework for systems development, but much of the book is given over to making the case for such a framework.

The first part of the book establishes that large-scale project failures are wasteful and can have dire consequences. There are numerous examples, including London Ambulance Service and the Bhopal gas disaster, culminating in the conclusion that management of most of these projects was incompetent and that it is virtually impossible to plan for projects that extend over years.

Bronzite contends that the methodologies used for systems development are not adequate, and he provides a selective, biased and slightly questionable view of available methodologies. He pays little attention to contemporaneous approaches, such as prototyping, object orientation, Rad/Jad and "formal methods". The book highlights the dangers of poorly managed systems development but does not reveal the power of these methodologies when used appropriately by competent people.

In the second part, Bronzite develops the theme that it is inherently virtually impossible to carry out accurate forecasting of resource allocation for large information technology projects. The budgeting for such projects, he suggests, is based on arbitrary limits set by senior non-IT executives. This results in the "over time, over budget" phenomenon often associated with IT projects.

Finally, Bronzite lays out generic guidelines for managing IT developments and advocates an incremental approach, making a series of small changes to existing systems. This seems to make sense, but given the discontinuities generated by IT developments, organisations that base their development strategies on this type of approach will be leap-frogged by competitors and new entrants.

The challenge is about managing in an age of discontinuity and uncertainty, and Bronzite does not get round to dealing with this fact. Perhaps if he looked at organisations that do manage competently, he might gain some insights for his next text.

Yasmin Merali is director, information systems research unit, Warwick Business School.