InformationWeek

The Secret CIO: The Year 2000 Advantage

A well-designed plan persuades the boss that adding resources can pay off

By Herbert W. Lovelace
Issue date: May 27, 1996

I am delighted with all the trade press coverage of the year 2000 crisis. It seems to me that the year 2000 turmoil, along with client-server and the Internet, will be enough to keep me and my fellow CIOs employed for the next few years. Meanwhile, consultants will get rich explaining the year 2000 problem and the proper methodology for fixing it. "Meth- odology" is a fancy word. It costs you a lot of money when a consultant uses it--even more money than when they use "paradigm shift." I won't use either.

The year 2000 is a problem because many systems have only two digits to store the year portion of each date. If you use such a system to check length of service and the resulting company benefits for someone who started in 1975 and will retire in 2001, the calculation will yield minus-74 years of service.

Recently, I got a call from Elizabeth, the friendly secretary that our president, Sir Marmalade, employs--no doubt to enhance his reputation as a nice guy who really doesn't know how Gornish, our CFO, and Kratmeyer, head of international operations, behave. "Mr. Whitestone would like to see you," Elizabeth said.

I probably should stop thinking of Philip T. Whitestone as Sir Marmalade--heaven help me if I ever slip and call him that in public--but his fondness for the sticky orange stuff is well-known. It seems that Phil had been reading about the year 2000 problem and wanted to know what we are doing ab out it. He seemed genuinely concerned--and for once took the time to listen.

I explained my plan, and--miracle of miracles, since he is not known for being decisive--he said it made a lot of sense and would support it. Because the plan meant we could save a lot of money on our conversion, he was one happy camper.

What I told Phil was that we have millions of lines of code that might contain dates that will blow our systems apart. I said we could spend a fortune searching for and fixing every error--but that if we were smart, we would concentrate instead on making sure we know when an error has occurred, determining if it was critical, and being able to fix the important ones quickly. I explained that I don't care about fixing two-digit date fields in order-entry screens--the customer service people will know that 05 stands for 2005--but I do care about them in interest calculations. We have a good track record of fixing systems when they go wrong, and if we invest in beefing up our responsiveness, w e can avoid shelling out some of the big bucks other companies are spending. As a bonus, our day-to-day service would improve.

Bomb Detection
Then Phil asked how I could be sure we would not be overwhelmed by the volume of problems. I said we were documenting what we knew about where the bombshells were, determining from our contracts the responsibilities of our vendors, and working on rapid response and test procedures. In addition, we were running dummy data though most of the popular systems while spot-checking for grievous errors. He smiled his winning smile and told me to proceed.

Having renewed my assessment of the wisdom of our president, I left his office, grinned at Elizabeth, and decided to drop in on my old buddy, Sid Gornish, who delights in making snide comments about me and my group. Some moments are just too good to enjoy alone and need to be shared.

Comments on this stor y?

InformationWeek http://techweb.cmp.com/iw