Lynn McGuire knows it's risky to make a big bet on WorldCom, with its sub-$2 stock price and shaky bond rating. But the CIO for Houston law firm Andrews & Kurth LLP also knows that telecom companies aren't delivering as many innovative ideas to help businesses as they once did. He believes that WorldCom's new Ethernet service is just what his firm needs--and AT&T, the firm's primary, and more stable, service provider, doesn't have anything similar.

During the telecom boom, technology executives such as McGuire came to expect a never-ending flood of big ideas for new services. Now, when a new idea does come along, it often engenders McGuire's concern about investing in a sorely needed service from a financially unstable company. "For our type of business, a service needs to be tried and true," he says. "If there's any risk associated with being on the cutting edge, there have to be significant economic advantages to it in order to justify the risk."

McGuire took the chance. Andrews & Kurth wanted to move its IT and administrative employees to a less-expensive remote office. That would be successful only with a high-speed connection that could be installed quickly. WorldCom promised to deliver within 30 days, so McGuire signed on for a 155-Mbps Ethernet link last month.

So far, it's working well for Andrews & Kurth--and perhaps even for WorldCom. McGuire says that in six months, when Andrews & Kurth's seven-hub, frame relay contract with AT&T runs out, the firm will consider replacing that network with WorldCom's Ethernet service.

It would be a nice ending for Andrews & Kurth and WorldCom. But the reality is that the telecom industry's historic spending boom--almost $400 billion in capital spending by the major U.S. service providers from 1997 until now--hasn't led to the great leap forward in data or voice communication that many had expected. Telecom's financial bust has been well-documented, with the bankruptcies of companies such as Global Crossing and plunging stock values among surviving businesses in the sector. For business-technology managers, the telecom boom-and-bust raised expectations, but carriers have been financially unable to deliver new services that solve even everyday problems, such as affordable broadband communication. That's forced many companies to throttle back on their plans for adopting promising new technologies, or in some cases to come up with alternatives on their own.

Houston law firm Andrews & Kurth signed up for a high-speed 155-Mbps Ethernet link from WorldCom despite the carrier's shaky financial footing. So far, the decision has worked out well, CIO McGuire says.

Telecom service providers admit it's a different world today. "The days are long gone when, just because a vendor has a switch that does something that we think is cool, we'll buy it," says Brian Brewer, chief marketing officer at WorldCom, the country's second-largest long-distance company and largest Internet backbone operator, which saw its stock fall from a high of $64.50 a share in 1999 to less than $1.24 in May. "We don't expect a return to the high capital expenditures of previous years." Reining in capital expenditures by spending only on equipment that will cut costs or result in new revenue "is the only way we can dig ourselves out of this mess," Brewer adds. WorldCom invested $11.5 billion in its network and new services in 2000. This year, that figure dropped to $4.9 billion. AT&T cut its spending from $10.5 billion in 2000 to $3.8 billion to $4.2 billion this year, Sprint's spending is dropping from $7.2 billion to $6.1 billion in the same period, and the major local service providers also cut their budgets.

Customers have adopted a more-conservative approach as well. Most business-technology managers still have a wish list of new services, but they need a powerful mix of cost-cutting and improved performance to take a chance in today's penny-pinching corporate-spending environment. "Making a bad decision is more costly than just going back to the basics," says Oray Auzenne, director of network technology services at Farmers Insurance Co. in Los Angeles. "When things were really growing in the industry, there was a lot more money available to invest in new technologies."

Auzenne's wish list includes voice-data integration as a way to reduce communications costs and add benefits, such as the ability to check voice mail via the Web. Farmers relies on AT&T for its data network, running both frame relay and asynchronous transfer mode at speeds ranging from 56 Kbps to 155 Mbps, for about 8,500 U.S. offices. "We want to leverage our network by routing both voice and data through the same infrastructure," Auzenne says. So far, none of the carriers has a clear advantage, and he's assessing options from several providers. AT&T is investing in voice over IP and in February introduced two managed voice-over-IP services.

Auzenne isn't sure which technology will end up being the best for routing voice and data traffic. It's a question that's remained unanswered since the late 1990s--ATM is still more reliable and expensive, while voice over IP costs less but isn't reliable enough. "It all depends on where the market goes and where you'll leverage the best cost versus performance, reliability, and scalability," Auzenne says. Until a clear winner emerges, Farmers will stay with separate networks.

Innovation isn't completely dead, from the telecom carriers or their customers. Service providers and equipment makers are focused on a few key areas: building local-access and global networks, establishing metropolitan area and long-distance Sonet and Ethernet services, and developing internal technologies that help them operate more efficiently. Still, customers say even those rollouts are slow, and the carriers are ignoring other areas that could deliver significant operational advantages to business, such as high-speed wireless data services and integrated voice and data communications. "The pace of innovation has definitely slowed, and I don't see the providers coming out with newer broadband services or the broadband wireless data services that they had talked about," says Michael Sherwood, CIO of the city of Oceanside, Calif.

During the mid-to late 1990s, the siren call of telecom technology held corporate America under its spell, entrancing companies with the promise of abundant bandwidth at steadily dropping prices. Racing to be the first to market and flush with borrowed cash, carriers unveiled a variety of features and services, such as ATM, virtual private networks, numerous flavors of Web and database hosting, low-speed wireless data, new features for frame relay, wireless LAN equipment, long-haul Ethernet, and the first voice-over-IP services.

The magic wore off in 2000, when it became clear that although data-service revenue would continue to rise, the growth would be much slower than expected. Service providers' meteoric growth in data-service revenue lost momentum, dipping from a 34% increase in 2000 to a 17% increase in 2001. An even weaker 11% growth is projected this year, according to U.S. Bancorp Piper Jaffray. The average per-bit cost for data transport fell 45% in 2001, the investment firm says, helping business customers but severely damaging carriers' bottom lines.

All that led to carriers pulling the plug on promising technology. Take Sprint's much heralded Integrated On-Demand Network, launched in 1998. The company had high hopes for providing advanced, integrated voice and data services over the network and had begun doing so. But the carrier decided to shut down the network in October 2001, resulting in a $2 billion charge against earnings and 6,000 layoffs. Sprint decided it couldn't get enough of a return on its $3 billion to $5 billion investment to justify the costs to operate it. "ION was very impressive and had lots of potential, but there wasn't enough financial support for it," says Patrick Wise, VP of E-commerce at Jacksonville, Fla., trucking company Landstar System Inc., which was evaluating the network shortly before Sprint pulled the plug. As Sprint's revenue has dropped, so has its capital budget--to $6.1 billion for all local, long-distance, and wireless operations this year, down from $9 billion in 2001.

The major telecom carriers say they haven't given up on innovation--they're just being more selective where they invest. For example, WorldCom says it's investing only in projects with near-term payoffs. In addition to its recently released Ethernet WAN service, WorldCom this month plans to launch a voice-over-IP service for businesses that it promises will deliver better reliability than consumer voice-over-IP services. "We believe voice over IP will be a significant product and revenue stream for the future in the business market, and we've invested heavily in that," WorldCom chief operating officer Ron Beaumont says.

Oceanside CIO Sherwood sees WorldCom's pursuit of the voice-over-IP market as a smart one, and it's an area in which carriers and equipment vendors have lagged in delivering on early hype. He's particularly disappointed that equipment manufacturers such as Cisco Systems, which provides Oceanside's voice-over-IP phone systems, have slowed their voice-over-IP development. He blames the slowdown for the fact that the city doesn't have voice and data integration for features such as unified messaging.

As with other providers, much of Sprint's investment will focus on technologies for its core network that deliver cost savings, including optical network gear that transmits greater amounts of data more efficiently and softswitches, which are cheaper than hardware switches both to buy and to maintain and update, says Fred Harris, the carrier's VP of research, architecture, and design. Sprint will increase its investment in the network once revenue increases, he adds, but for now, one primary area of focus is secure IP services.

AT&T doesn't spend much on what it sees as visionary technology, meaning products that will take more than a couple of years to become mainstream. Instead, it favors a course it calls "relentless incrementalism," says Dale McHenry, AT&T's VP of high-speed packet services. That strategy has protected AT&T from misguided investments in cutting-edge technologies that never lived up to their potential, he says. AT&T has budgeted half as much for capital spending this year as it did last year; it plans to spend the bulk of its $3.8 billion to $4.2 billion capital budget on expanding network capacity and building overseas networks. It's also investing in local networks, VPN services, digital subscriber line offerings, voice over IP, network-based firewalls, and managed-service options, as well as in optical switching and transmission equipment for its U.S. data backbone network.

Big telecom customers still can get carriers to jump, but even they can't always get exactly what they want. The U.S. Department of Energy and NASA teamed to buy an extremely high-speed network linking about 30 research sites. Sprint had been their ATM network provider for five years when the agencies went looking for proposals in 2000, but Qwest Communications International Inc. won the latest contract, in large part because of its promise to develop 40-Gbps access. That's a service no carrier, including Qwest, offered at the time. "Qwest promised the capabilities, and we have confidence in its proposal," says Pat Gary, network projects leader for NASA's Goddard Space Flight Center. By agreeing to innovate, Qwest won a $50 million deal.

But NASA and its partner agencies won't stop there. They want access speeds of 80 Gbps to 160 Gbps in the future. If Qwest or other carriers are unable to deliver those speeds when the agencies need them, they could buy dark fiber and install and manage their own equipment to combine multiple circuits to get the desired speed--a warning to carriers that if they don't innovate, customers will. "It appears that some of the technology challenges we face are solvable," Gary says.

Kaiser Permanente agrees. Its 150-line T1 private line network wasn't fast enough for the data-intensive medical imaging applications the Oakland, Calif., health-care organization runs. So the company decided to build its own Gigabit Ethernet network using leased fiber-optic circuits and its own equipment, says senior programmer Joseph L'Italien. "What I'm trying to do is use our existing equipment to provide better connectivity for our users," he says.

If businesses want to keep innovation alive, that's exactly the approach they should take while the telecom industry is in disarray, says Russ McGuire, chief strategy officer at consulting company TeleChoice Inc. That means pushing the limits in their own networks as much as financially and practically possible.

One of the most common complaints is the lack of fast access. For example, service providers haven't built fiber-optic networks fast enough to keep pace with Oceanside's need for broadband data services to connect central computer networks to remote sites. "We don't have our own fiber infrastructure in place between sites, and that's what we need for the services like Gigabit Ethernet that we want," Sherwood says. No local providers have the fiber in place, nor can they tell him when they might be able to install it.

During the past few years, business customers had come to believe that cheap bandwidth would be readily available. That hasn't happened, says Matt Kesner, chief technology officer at Fenwick & West LLP, a law firm in Palo Alto, Calif. The carriers' backbone networks contain a glut of bandwidth after the massive buildup of recent years, but there aren't enough high-speed access lines--the "last mile" connection to the customer--to use that capacity.

A number of startups, including Broadband Office, Cogent, and Yipes, aimed to offer super-high-speed access and backbone circuits to heavy-duty customers. But Broadband Office went out of business, and Yipes filed for bankruptcy in March. Many of the service providers and equipment companies still in business are just catching up to what startups had been offering, Kesner says. His primary local provider, SBC Communications Inc., still doesn't have metropolitan area Ethernet services, which provide native Ethernet transport at high speeds, available in all the areas around Silicon Valley that he needs. Where services are available, they're pricey. "We have big pipes, but we pay a lot of money for them," he says. Even for home access, Kesner says, a quarter of his staffers live in areas where they can't get DSL.

In addition to the price, nationwide companies complain they can't get the coverage they need. Newer services, such as Ethernet metropolitan area network and wide area network offerings, are either too expensive or don't have the national reach and uniformity, says Bill Baumbach, VP and CTO of Bear Creek Corp., the parent company of gourmet food retailer Harry and David and other retail businesses. Bear Creek, which has more than 100 Harry and David stores across the country, hasn't found an affordable national data service managed by one provider, Baumbach says.

In addition, Bear Creek's cyclical business means it overpays for data capacity that's idle most of the time, because the carriers won't let the company adjust the data bandwidth of their circuits up and down in response to traffic fluctuations. "We are forced to purchase for our maximum usage and don't have the flexibility to quickly adjust capacity to our needs," he says.

Landstar has found similar problems linking its sites. The company needs to communicate with about 8,500 truck drivers criss-crossing the country at any given time, and at least 50 or 60 small offices or telecommuters' homes. "I have people all over the country, and that's a huge problem," VP Wise says. Connecting each of those offices requires contacting different carriers and figuring out which of the voice and data services they need are available in a given area. "Lots of times, the services aren't available," he says.

Mobile data services are Landstar's biggest need, Wise says. Truckers who need updated shipping information wrestle with an unwieldy mix of older technologies, including dial-up connections and dial-up accounts from Internet service providers. The most advanced use mobile data applications based on the Wireless Application Protocol standard on their cell phones. But many others just get by with regular phone calls, often relying on a network of family members to pass along the latest information, says Wise, who's also investigating satellite-based data services for the fleet. "Getting text to them has just been a horrible nightmare," he says.

Wise says he's looking forward to the launch of nationwide data services on third-generation cellular networks and to more widespread introduction of wireless LAN services running on 802.11b technology deployed in public areas, but he doesn't expect either of those to be available in the near future. AT&T says it's exploring those areas.

Customers frustrated at the slowdown in telecom innovation shouldn't expect changes anytime soon. Depending on many factors, the telecom sector won't fully bounce back from its troubles until at least 2004, says Ron Kline, an analyst with RHK Inc. Combined capital expenditures will hover around $46 billion to $51 billion this year and next year, Kline predicts.

But slower rollouts of new services might not be the worst result of the telecom meltdown. A bigger concern is whether the industry's problems are a drag on revolutionary thinking. That would restrict technological progress to incremental changes that won't provide any fundamental advantages over what's available today, says Jeremy George, director of the advanced networking group in IT services at Yale University. Yale is testing voice over IP, converged voice and data communications, and PC-based "softphones" in small groups of fewer than 100 people.

Rather than being seen as cheaper replacements for regular phones, the new technology should get people thinking about converged networks as creating new ways of communicating. But that's not happening in many places within business or academia. "People are less likely to take risks and try new technology," George says.

Yet George sees hope. While excitement may have cooled at telecom companies--and some of their customers--for new services, today's 20-year-olds won't be content with the status quo. "In five to seven years, when they get positions of power, they'll begin to implement this technology" in ways that older people can't foresee, he says. That just might bring the telecom revolution people have been waiting for.--With Robin Gareiss