Developing a Plan
After attending the intranets conference, we seriously considered the purchase of an out-of-the-box portal solution. At $35,000, it wasn't too expensive, and it seemed to offer many of the features we wanted within our intranet, including the capacity to support approximately 1,000 users. But before we made the purchase, we thought we should attempt to create our own site and see where we landed with it, since it was quite possible that anything we created could be incorporated in the purchased package.
A Doodle Becomes a Mascot
Many very interesting ideas were introduced at the intranets conference, and we had all agreed that we wanted to include some of them in our development project. Something another company had done was the introduction of a mascot. We decided to give our intranet a real name and personality, instead of always just calling it "the intranet." (Some people got "intranet" and "Internet" confused anyway, so we needed something completely different when referring to our internal Web site.)
Since our company name is Midland and this was basically a "com" project, we went with the name Midcom. But Midcom needed a personality, too, and thus the "Midcom Guy" was born (see Figure 2). He started out as some silly doodles using the company logo, but he was quickly adopted. The Midcom Guy is now on every page of the site, with various "looks" depending on the theme of the page. For example, in the Cafeteria section, he's wearing a chef's hat and holding a spoon.
He may not look like much, but the Midcom Guy gave the intranet a real "personality."
Improved Navigation
Since navigation had been so poor in the original intranet, we wanted to avoid making the same mistake with Midcom. From the beginning, we established a rule—the Back button should never be needed to move between pages. If someone must use that button to navigate, we had failed. We also wanted a menu that would expand as we added more content. And we tried to stick to the "three-click" rule, meaning that it shouldn't take more than three clicks to reach a destination within a Web site. That doesn't mean that you should be able to access everything on a site in three clicks—just that you should be able to find what you want in three clicks. If you want a telephone number, for example, you shouldn't have to click eight times to find it.
Developing a Style
The original site had very little consistency in colors, graphics, or fonts. An informal "style guide" was established, with Arial (Helvetica/sans serif) as our font of choice, and a blue and a gold from the Web-safe color palette of 216 colors. All backgrounds would be white, all text would be black and within certain sizes, and so on. It helped that only two of us were doing the actual HTML pages, so we were able to maintain this consistency using cascading style sheets. We later went back and applied these rules to all previously created databases and calendars that we planned to move from the old intranet to Midcom. Finally, we wanted to make Midcom a fun and cool place to go, and hoped to incorporate some of the "portal" look and feel.
Unfortunately, our network requires two logins to reach the intranet: You first must log into the network itself, and then you must log into the intranet in a browser window. This problem wasn't going away anytime soon (and is still with us, I'm sorry to report), so we had to devise a way to draw people into Midcom on the login page. Even though the browser is launched whenever a user signs onto the network, and the home page is the Midcom login, we still didn't feel this was enough. Therefore, we agreed that news headline–type "teasers" would be listed on this page, with instructions telling users they must login into the Intranet site to "view current news."
No Pressure, No Diamonds
All of this had to be done by Friday, May 5, 2000. It became known as the Live by 5-5 Campaign. We had approximately four months to deploy Midcom, but as I mentioned earlier, this wasn't a full-time role for most of us, and fitting all those tasks into a 40-hour work week usually meant it was 60+ hours a week instead. We decided to meet weekly to stay on top of everyone's progress, although email updates happened almost daily.
In addition to the project deadline, we also had another deadline: The lead Lotus developer was several months pregnant, and her baby was due around the launch date.