ERCB: Software Runaways

[an error occurred while processing this directive]

Vital Statistics

Title	Software Runaways: Lessons Learned from Massive Software Project Failures
Author	Robert L. Glass
Publisher	Prentice Hall Professional Books Upper Saddle River, New Jersey http://www.prenhall.com/
Copyright	1998
ISBN	0-13-673443-X
Pages	259
Price	$32.00

Software Runaways

One of the thoughts that sprang to my mind when I read Software Runaways by Robert Glass is that it is a good thing for the book, the author, and the publishing house that the world operates by Adam Smith's theories, where value is determined by supply and demand, instead of Karl Marx's, where value is determined by the work done. Perhaps that is a little too harsh, as Glass surely had to spend considerable time and energy compiling his articles, asking for permission to print, and lining up people to write their views on sundry topics. Nevertheless, it should be pointed out that a good portion of the book (half?) consists of articles culled from print media of widely ranging technicality (from IEEE Software to Computerworld to the Wall Street Journal). On the other hand, most of the book, including Glass's own writing, was very readable and sometimes humorous.

At the beginning, Glass lists 6 major causes for failure along with a few pointers to potentially useful sources, such as a KPMG study published in Software World, vol. 26, no. 3 titled "Runaway Projects -- Causes and Effects". The six causes he lists are:

Project Objectives Not Fully Specified
Bad Planning and Estimation
Technology New to the Organization
Inadequate/No Project Management Methodology
Insufficient Senior Staff on the Team
Poor Performance by Suppliers of Hardware/Software

Certainly this list should inspire discussion. For instance, the case could be made that very few long projects could have their project objectives fully specified at the beginning since business requirements change fairly rapidly in most fields. Suffice it to say that the introduction was interesting enough to induce me to buy the book.

The great bulk of the book comes in the middle section, which consists of a collection of "war stories." The reason why I put the term in quotes is because the majority of the pieces were written by third party observers, not actual participants of the development efforts themselves. The pieces themselves were written with varying quality, of varying depth, and for different audiences.

As a broad overview of major software disasters, Software Runaways is quite complete, but I often found myself frustrated. Glass tended to be impartial, allowing more than one viewpoint in on the more acrimonious failures, and declining to offer judgment on most of the case studies (though a few of the writers did so themselves). Perhaps a smaller collection of case studies researched more in depth would have been more profitable.

The book ends in a section titled "Software Runaway Remedies". Its format is very much like the introduction, in that pointers to other source material are offered along with some discussion of their conclusions. There are some thought-provoking and worthwhile observations made -- for example, the book prompted me to ask if my organization did postmortems on projects. The length of the final section, though, was shorter than expected. In summary, this book is a good read, and considering the paucity of useful material available in the trade press on software failures, may be a good buy.

-- Richard Tung (rtung@godzilla.cs.nwu.edu)