Global CIO: IT Project Triage (And Why Skippy Must Die)

You have a problem.

Your project budget has been decimated. The suits are under serious budget pressure and are mouthing expressions like "shared pain," which is never what you want to hear. So you will have to decide what lives and what dies. Further, some of your best people are on projects that will never see sunrise. Did I mention that there are some Sacred Cows out there protected by their Godfathers, but which should logically die? Can you figure out which Godfathers themselves are on their way out?

This isn't about technology; it's about management. And you need some help to plow through this mess and get to a point where you can do the fun part -- which is showering money on really sexy things that will wow Mahogany Row and drive the business forward. But now is not the time.

Some of these projects are "strategically important" but might not survive the bloodletting -- is there a way you can hide them? Some of these projects have so much management attention that you dare not kill them, but they should mercifully be put out of their misery, either because they're never going to work or the real cost is three times what anyone thought. Other projects made sense at the time but don't now. Want to take that Big Write-off now? Not such a good time, is it?

Want to play company politics? Very risky. Ignore politics? More risky. This isn't the time to bet your job. Think of the entire project pool as one large Dungeons & Dragons game.

Howard's Rules

• Find a common enemy. Maybe it's the economy. Maybe your company is at a crossroads. But use the common enemy argument to kill the obvious losers. Kill any project where the return on investment -- and you know how to fudge those numbers -- is more than two years out. Kill projects where the resultant savings/benefit cuts over multiple cost centers. Kill projects whose justification is flimsy, like it will save everyone 6.3 minutes per week, which will never happen. Some projects are like statues in the park: They started with great hoopla but today no one quite remembers the subjects.

• Protect your best people. Move them into Safe Harbors, projects that can't be killed, even if those projects aren't quite as much fun or challenging. A great programmer is worth 10 average ones. A great project manager is the difference between "on time, on budget" and Excuse City. Yes, you may lose a few people, but you'll live to fight another day.

• Protect the projects that keep the lights on and will carry you to a better day. There is always a tendency to let them slide and put off upgrades until "next year" -- but next year may be worse.

• Find some projects to throw under the bus. You must show that you are a Team Player, so know what you want to kill and why. Smart CIOs will start to move their deadwood to those projects now, so that when they get killed, the people you would like to go will go with them.

• Get the operating divisions to kick in some of their budget to the Sacred Cows. That will force them to choose.

• Keep one or two Knock Your Socks Off projects. You need to retain a little sex appeal to give hope to the superstars. Do as little as possible as loudly as possible.

• Pass out enough sugar to international so they don't feel completely neglected.

• Combine projects where possible.

• Realize that what you're buying is Time. You just don't have the budget you thought you did. Cutting each project by 15% doesn't really do it if your budget is reduced by 15% -- it should, but it doesn't. Some projects must be cut to zero. And they must be cut right now.