Guest column: Net Prophet

Next week is the last week of December. The hatches have been battened. The patches have been applied. The ammo has been stockpiled. Now there's nothing more to do but wait for year 2000 and see if the world ends.

But we know the world won't end, and when the year-2000 problem is but a memory, we are all going to look back on the whole experience fondly.

We'll view it as a glitch from a kinder, gentler era, when we started to deal with Internet-commerce and electronic-business problems.

The world will not have ended because of year-2000 problems. This is a world that uses Microsoft Windows everyday, so it's not like we're used to glitch-free computing -- but maybe we will have learned a few things.

Market research company Gartner Group has issued a doom-and-gloom report claiming that 75 percent of e-business projects will be deemed failures through the year 2000. Now, the definition of a "failure" is slippery, but that is a prediction that should send chills down the spine of e-business strategists everywhere. Even more disturbing, Gartner Group predicts that 5 percent of all enterprise systems will fail through the year 2001.

Gartner Group's predictions might seem a bit alarmist at first, but consider recent high-profile problems experienced by some of the corporate world's biggest players.

Hershey's took a hit this Halloween thanks to problems with an enterprise resource planning (ERP) implementation. These problems translated to a 19 percent drop in third-quarter profits and a 12 percent drop in sales.

Whirlpool's push for a Labor Day buying weekend launch of its SAP R/3 implementation resulted in backed-up inventories and delayed shipments.

W.L. Gore & Associates -- the makers of Gore-Tex -- is actually suing PeopleSoft and megaconsultants Deloitte & Touche for a botched ERP implementation.

Finally, consider Toys R Us and its Web site relaunch for the holiday shopping season. The company somehow managed to grossly underestimate its I-commerce capacity needs, with the end result being locked-out customers -- not a pretty picture for a retailer at holiday time.

E-business looks pretty darn swell on paper. Unfortunately, e-business solutions don't work on paper -- they work on computers.

Now, mission-critical systems aren't anything new, but those systems have traditionally been fairly discrete. The sheer pervasiveness of e-business makes companies vulnerable to devastating glitches.

So what can we learn from our year-2000 experience that can serve as our guide during the inevitable e-business and I-commerce failures?

First, we have to consider how problems might be avoided. I'm not one of those people who lambastes the programmers for using only two-digit dates. That just ignores the economic realities of the time.

I'll even forgive the lack of an upgrade plan because nobody seriously thought we'd still be using so many of these systems today. But, once bitten, twice shy.

Now we have to look at all our various ERP, e-business, and I-commerce applications with a critical eye toward problems. Every software project makes compromises. But, in a post-year-2000-problem era, there is no excuse for not having a plan for when those compromises become untenable.

There is currently a mad rush to be first. Time to market is the driving force of application development because historically, first movers have gained market advantage.

But we will soon see a time when the pressures of being to market first collide with the reality of what can be coded. The result will be some spectacular, high-profile explosions.

The world's year-2000 problems should also have acquainted us with our friend, Mr. Contingency Plan.

Perhaps it's obvious, but you do not want to be hammering out a contingency plan when you find yourself in the middle of a site meltdown.

Every company that does serious I-commerce business needs a disaster plan. It should include key items such as who to call and when they can get there.

Trust me, nothing makes you look worse in the eyes of the press -- and we will be reporting on your disaster -- than when it seems like you don't know how to handle a big system failure.

IT project failures aren't anything new. However, the Internet is one big glass house where the world can see into your failures.

Being an e-business that takes advantage of the Net also means you're subject to the scrutiny of the world.

No amount of testing will uncover all the problems. That's a fact of life.

Only the real world will expose certain flaws and bugs in a system.

And, when those problems rear their heads, you'd better have a plan for how to deal with them.

So, let's consider our year-2000 problems and experiences a friendly reminder.

Sean M. Dugan is senior research editor in the InfoWorld Test Centre

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about Critical SystemsDeloitte & ToucheGartnerGartnerMicrosoftPeopleSoftSAP AustraliaToys R Us

Show Comments