My IT career began just as the Internet boom was ramping up. Not surprisingly, when my company's management team noticed our competitors moving onto the Web, it decided we needed to be there too. After several meetings, all the various divisions and departments came to an agreement on a simple promotional site. A couple of weeks later, we were online. Everything would have been peachy, except for "Stanley."
One day I found a yellow sticky note on my monitor: "Come see me ASAP -- Stanley." Stanley was one of the more clueless managers on the Web Design Committee, and he gave me a long, angry lecture about how the size of the corporate logo on our Web pages was unacceptably large. He demonstrated the problem by measuring the logo on his computer monitor with a ruler and comparing it to the height of the logo on the company letterhead. He insisted that the online logo had to be the same size as the one on the letterhead, and he ordered me to resolve this issue immediately. So I sat down at his PC and changed the display resolution from 640 x 480 to 800 x 600. "Now, that's how you should have done it in the first place," he snapped.
A year passed, and suddenly Stanley started using the word "portal" regularly. Apparently he'd been invited to some product demo. "You're going to build a portal," he told me one morning, "using tools from Construct-a-Portal [not their real name]. I want live, real-time financial reports and metrics available to everyone in the accounting department via the portal." He wanted us to go live within the month; the demo he'd seen showed how easy it would be.
We ordered a new server and a license for Construct-a-Portal. Fifty grand went poof. A week later we advised Stanley that the software appeared to be seriously buggy and we needed some training. Off I went to Construct-a-Portal's training center for a four-day crash course. Seven grand went poof. When I explained our project to the course instructor, he matter-of-factly informed me that we were using the wrong product. Posting live metrics was not possible with Construct-a-Portal's product; doing so would require advanced workarounds.
Even so, we managed to get 50 reports online. Now the only issue was navigation. Stanley insisted that all the reports had to be available on one screen, with tabs across the top. We explained that a screen with 50 tabs across the top was a bad idea. But he insisted, so we created the despised "Tab Window," with four rows of tabs almost filling the screen. If someone wanted another report, we'd add another tab. It was the ugliest thing you ever saw.
Four months past deadline, we went live. The first month's logs show that Stanley accessed the system twice, and some member of the financial group logged on once. The following month nobody logged on at all. Eight months later I got a call from a new IT hire wondering if he could use the server for something else; no one had accessed the portal for months.
This project took five months and cost $150,000 to $200,000 in hardware, software, and man hours. The only thing of value that came out of it was a running gag: "When in doubt, add a tab."
The lesson? Don't let impressionable managers anywhere near a product demo.