Columns

We ought to plan to fail, right?

• By Michael Alexander
• November 1, 2004

As Patricia Keefe writes in "Oops! Ford and Oracles mega-software project crumbles," the bigger the project, the more likely it will fall apart before it’s done.

Ford had high hopes for Everest. When the project was announced, Ford said it would create a standardized, Web-based, system that would be accessible 24 by 7, dramatically increased the speed at which it and its suppliers communicated, and “e-enabled” employees with better decision tools. The company reportedly dumped more than $200 million over five years into the project before it turned off the switch. That’s a conservative estimate, by the way. Some analysts calculate that Ford reached twice as deep into its pockets.

Ford tried to do the right thing but went about it in the wrong way according to one analyst. Ford and Oracle declined to talk about it, but some observers say Everest was ill conceived and too unwieldy to integrate. The coup de grace was the project failed the “so what” test with Ford’s suppliers who had to step through a portal and five screens only to find they were unable to find the data they needed. No wonder some started calling the project “Neverest.”

Naturally, most enterprises expect to succeed when they launch mega-software projects. The odds are in their favor, but just barely. About 40 percent of the time they fail, say the people who keep track of these things.

So does it make sense to plan to fail as much as it does to succeed? Yeah, it does. Keep reading Keefe’s piece and you’ll find out what the experts have to say about recognizing when it’s time to pull the plug.