Hello? Hello? Can you hear me now? The telecom sector seems badly disconnected. Analysis reports state that over two trillion dollars' worth of its market value has evaporated in less than 30 months. The high-flying, high-tech visionaries of the high-bandwidth future-Global Crossing, Covad, Williams, XO, Teligent, et al-have vanished into bankruptcy or liquidation. The AT&Ts, WorldComs, Qwests and Sprints, as well as their counterparts overseas, have seen their bold ambitions for growth in billion-dollar gambits such as the third-generation wireless standard turn into mad scrambles for survival. A few dishonest telco execs may even be going to jail.

There are many good reasons for this sorry state beyond corrupt accounting. Here's one of the best: America's telecom companies are lousy innovators. They've been more comfortable experimenting with their debt financing than experimenting with new service offerings. Even worse, they appear married to hopelessly outdated definitions of innovation. Telecom equipment suppliers do a sensational job of creating new bandwidth and extending physical networks-sensational, but counterproductive, given the current capacity glut. Does new capacity that nobody is willing to use or able to afford truly count as innovative?

Innovation isn't what companies do; it's what customers adopt. In fact, the telecom sector remains a fabulous market for innovative uses of bandwidth. But innovation shouldn't mean getting people to use more bandwidth; it should be about getting people to change their bandwidth behaviors.

Clever features like NTT DoCoMo's i-mode Internet services and handheld wireless games (see The Wireless Arcade, TR July/August 2002) can be tremendously appealing and even successful. There is also no shortage of marketing campaigns-most notably AT&T's mlife-that continue to promote telecommunications as a lifestyle revolution. But this revolution has created more casualties than converts. The multibillion-dollar bids to create bandwidth-based lifestyles via the cell phone failed.