Avoiding IT Disasters

by Don Estes

This seems to be the year for disasters. In the last three months, in addition to the September 11 disaster, I've seem three very large IT projects cancelled after the expenditure of considerable sums of money. One was in excess of US $50 million -- an ambitious Web project whose managers did not think it necessary to include performance modeling in the design. A second was more than $100 million, which was an attempt to rewrite a very complex 30-year-old system. And the third was in the hundreds of millions, and involved attempting to convert to a package "with some changes." All were DOA. Needless to say, careers have been cut short as a result.