I listened in to a pretty frank discussion between a few systems integrators recently. The conversation went along these lines
"The 80s - the custom development arena - glad all those budget overruns and scope changes and delays are long gone"
"The 90s - we just tried to slam client/server ERP software in - there was no re-engineering or focus on benefit stream and that explains the huge overruns"
"This decade - we are so much smarter - we know where the payback
points are. Net of payback, project costs are so much more manageable"
I do not know, guys. No SI will ever know an industry or pay back areas better than a client team. Not sure SIs can take too much credit for the benefit stream. What they can control is the scope and cost equation - and not sure the industry has really learned to manage large projects well even today. This oft-quoted Standish analysis is a good reminder of the industry track record around large projects.
I have always wondered whether the industry would benefit from hiring program managers from the Bechtels and Fluors of the world - the combination of project management rigor and the cynicism that comes with planning for uncertain physical logistics. It is the latter the SI world lacks as it continues to run projects primarily by GANTT charts.
