• E-mail this page
• |  Print this page
• | 
• | 
• | 

Tech Disasters Are Just Waiting To Strike Your Organization

"Can't anyone here play this game?" That's what Casey Stengel famously asked about his hapless 1962 New York Mets. The same question might be posed to those running big-time corporate and government IT projects, given the frequency with which such efforts end in tears. In Monday's InformationWeek, we put the spotlight on eight infamous tech disasters that resulted in major financial losses, damaged careers, and abrupt interruptions in business operations for those involved. Things have obviously gotten better for the Mets lately. But the winning percentage for grandiose tech initiatives remains anemic. What the heck is going on here?

The consensus is that in most cases overly complex IT makeovers are doomed to fail because their success depends on too many unpredictable variables falling nicely into place. That rarely happens in the real world. When one link in the chain snaps, problems mount in a hurry. At Nielsen Media Research, to cite one of our examples, tech execs tried to rush a complete rewrite of the system that drives the company's billion-dollar research business. But coding errors led to problems, which led to other problems, which led to the fact that Nielsen is still trying to get it right after almost 10 years.

Change management failure is also a common theme that runs through our horror stories. A number of the featured players were in the midst of application upgrades, hardware refreshes, course corrections in business technology strategy, or other disruptive about-faces. We came across one publicly traded firm that somehow unplugged its accounting system for three months during a merger, and an insurance company that lost $20 million buying, and then selling, an offshore outsourcing unit within the space of a year.

A 2001 study by IDC showed that 78% of all IT downtime is related to change.

There's more. And it's enough to make some of the more conservative CIOs out there batten down the IT hatches for good. But the "ain't broke-don't fix" strategy for survival is, in fact, a path to extinction. Information technology is constantly evolving and businesses and governments need to keep up or lose their edge. The ostrich approach won't work in today's business-at-the-speed-of-electrons age.

What's needed are more meticulous business case analyses, sound strategies for change management, and workable back-up plans if the big one happens. Speaking of the latter, the IRS earlier this year tried to implement a new anti-fraud system in time for the 2006 tax season. Bugs kept it from going live, which was too bad because the tax agency had already trashed the older version. You and I are likely to be paying for that one for the next few years.

So the IRS joined our list of organizations that, through lack of communication, poor planning, outsized ambition, or just plain incompetence, have suffered through embarrassing IT disasters. But it's more than just human fallibility at play here. Increasingly, we're delegating to computers more responsibility for carrying out the day-to-day functions that keep our businesses, governments, schools--indeed our whole civilization--running smoothly. But Murphy's Law trumps even the most sophisticated forms of automation every time. What can go wrong is still going wrong. And as IT takes on a more central role in, well, everything, things are going wrong on an ever grander scale when systems do pack it in.

What's the upshot for CIOs at businesses that depend on technology for their very survival? In deference to the World Series, think of it this way: With so much at stake, you want to take your playbook from the 2006 New York Mets, not the 1962 version.

« Daily Podcast For Friday, Oct. 13, 2006 | Main | Printer Ink--The New Black Gold »