Beware of business IT horrors 

In this issue, my colleague Chad Dickerson unveils his “Top 20 IT Mistakes” based on some unhappy lessons he and others have learned. This list is so much fun that I couldn’t resist adding my own: five common mistakes that business people make when dealing with IT. Each of these derives from bitter experience. And in three of the cases, I was personally responsible for the bad judgment (please don’t ask which three!).

Underinvesting in New Projects Nothing hurts an innovative technology product more than skimping on its budget. In the early days of the Web, I was part of a team that launched a novel and clever online data service. But we pushed the envelope of then-current technology using an unproven platform running on minimal hardware with about one-third the needed engineers. We did finish the darn thing. Sometimes it even worked.

Consultancy Addiction Years later, I worked at a company that was developing a new line of business. This was during the dot-com boom, when you couldn’t hire programmers for love or money. They were all working as $300-an-hour consultants. So my company hired a friendly consultancy to help us ramp up. I remember crowded planning meetings where the billable hours probably exceeded my annual salary. Nothing was ever built.

Not Sweating the Small Stuff Business and IT people alike often underestimate the amount of upkeep that systems require. And as your project list grows, these seemingly small maintenance functions tend to get pushed off in favor of projects that boost revenue or make dramatic fixes. At one company, the small stuff was pushed off for so long that eventually the entire system ground down to a mysterious, molasses-like performance that took months to diagnose and cure.

The Eternal Two Weeks Software projects often go through a phase when they may appear near completion but in fact hide many problems. You’ll know this when the IT folks keep talking about a project being “two weeks” from launch — or two months, for big projects — yet the two weeks never end. I remember one project that remained in the two-week stage for nearly a year before it was granted a mercy killing.

Champagne Delusions The error here is assuming that because a tech product is launched and the champagne’s been opened, the work is done. Such products are typically about half finished at launch, despite everyone’s best intentions. So treat the champagne as celebrating the front nine; the back nine are yet to be played.

Got a business/IT horror story of your own? Send it in. Names will be omitted to protect the guilty!