Let's look at Motorola. This was a technology company; it did integrate backwards into chips--remember that all those early Apple computers ran on 68000s. It was a supplier to the carriers and got out of spectrum so it wouldn't be seen as a competitor to the AT&Ts, Verizons, and Sprints of the world.

It was the carriers that hobbled Motorola, because there was more capability in the phones than the carriers would allow. Along came Apple, violating the conventional wisdom by making its smartphone pitch directly to the consumer. And suddenly Motorola had an obsolete product line and a weakened hand. It lost first to Nokia, then to RIM, then to Apple. Cell phones became fashion purchases, and Motorola found itself in the backwater. What should MOTO have done? Gone into the carrier business and competed with its existing customers? Pushed the idea of the smartphone directly to the consumer?

Nortel owned the high end of switches; its customers were the world's telecom authorities. It was a major factor in routers as well. What went wrong? This was not a flagrantly mismanaged company, was it?

In fact, it was. Nortel went on a spending spree the technology world had rarely seen before, acquiring Alteon, Bay Networks, Qtera, etc. These mild-mannered Canadians had a dream: to be the next Cisco. If Cisco got credit every time it made a multibillion-dollar acquisition, whether accretive or not, couldn't Nortel do the same?

Yes, for a while. Then the telecom boom ended, the 300 CLECs (competitive local exchange carriers) found their funding cut off, and all of a sudden there was no market and these expensive acquisitions were hemorrhaging cash. Not that Alcatel/Lucent was doing any better.

So here we have three poster children--all technology-driven, all with a "franchise," all with happy customers, great brand names, and strong balance sheets. And all close to pushing up daisies.

When Intel was having its problems 20 years ago, Gordon Moore and Bob Noyce asked themselves what a new CEO would do if he inherited the paper-thin profit margins that memory chips were generating for their firm. Their answer: A new CEO would immediately get out of the memory business. So they did. Fortunately, there was a great new opportunity on the horizon: personal computers. If you're going to jump out of an airplane, it helps to have a parachute.

IBM also flirted with its own death spiral. It was as hooked on mainframe hardware, mainframe software, and mainframe profits as Kodak was on film. Having realized that those glory days were never to return, IBM rebuilt itself on a platform of software and professional services, jettisoned hordes of employees, and got out of PCs because it saw that it was never going to make money in an increasingly commoditized business.

The first job of leadership is to be stark realists. IBM and Intel were. Kodak, Motorola, and Nortel were not.

A few columns ago, I decried One Generation Companies such as Novell, 3Com, and Palm, which all flowered but eventually burned out. Kodak, Motorola, and Nortel are different. They're 75 to 100 years old--four- and five-generation companies. And even though all three were well run for much of their existence, in the end they weren't.

They had good, not great, leaders. They recognized change and adapted, but not quickly enough, not violently enough. Hewlett-Packard is having its own Come To Jesus Meeting even as we speak.

In the end, these three icons, Kodak, Motorola, and Nortel, thought they were managing risk. In the end, it didn't matter.

Howard Anderson, founder of Yankee Group and co-founder of Battery Ventures, is currently the William Porter Professor of Entrepreneurship at MIT. He can be reached at [email protected].

For more Global CIO perspectives, check out Global CIO.

For the 15th consecutive year, InformationWeek is conducting its U.S. IT Salary Survey. Upon completion of the survey, you will be eligible to enter a contest for prizes including a Bravia HDTV or iPad 2, and get a link to download our report once it is published. Take the survey now. Survey ends Jan. 20.