Net innovation gets squeezed

Part One: Network industry rearranged
Part Two: Slim budgets test users' creativity

Beyond today's layoff news and this year's shrinking network budgets, the ongoing economic turmoil could have a long-term impact on the lifeblood of the technology industry: innovation.

After all, layoffs, research and development cutbacks, and a lack of start-up companies mean fewer people are working to solve network problems, and fewer ideas will ever see the light of day. On the other hand, observers say what is really emerging is a more Darwinian environment than we have had in recent years, where only the strong ideas survive.

"We were in an era where almost any idea could get funding . . .," says Bill Leighton, vice president of network and services development at AT&T Labs. "So many new ideas were generated [that] people like myself had to spend a lot of time sorting it out."

This, the final part of Network World's series about the economy's effects on the industry, focuses on how innovators plan to make the most of a bad situation.

"It's inevitable that we'll go forward . . . with technologies like wireless Internet, optical and mass-market broadband," says Larry Smarr, director of the California Institute for Telecommunications and Information Technology (CalIT2) and a member of President Bush's IT advisory council. "People won't stop innovating, but it's going to be short rations for a while."

That's partly because private investment in network start-ups - a major engine of innovation - evaporated this year. The second quarter reported the lowest amount of seed- and first-round financing deals in the industry since 1998, according to PricewaterhouseCoopers' MoneyTree survey. The survey lists 116 seed- and first-round deals in the second quarter, down 80% from the peak in the first quarter of 2000. At $688 million, the total financing in the spring of this year was one-tenth the amount spent in early 2000 (see chart, right).

Acquisitions of start-ups are also down. For example, Cisco - which gained a reputation for acquiring innovation rather than developing it in-house - bought just one company in the first half of this year, compared with more than 20 last year. And IPOs have plummeted in the technology sector, from more than 300 in each of the last two years to just 10 in the first half this year, according to the Web site IPO Monitor.

For corporate buyers, this trend means start-ups will take longer to deliver products because they will have less money to spend on people-intensive development efforts. However, start-ups that deliver products should be healthier because venture capitalists are doing more due diligence before making investments.

"You will see stronger businesses because companies won't get funding unless they have a compelling business plan and a revenue model that shows they can make it," says Tracy Lefteroff, global managing partner of the venture capital practice at PricewaterhouseCoopers.

PricewaterhouseCoopers executives say they won't worry about innovation in the network industry unless early-stage investments drop below 1997 levels and stay there for more than a year. Their overall view remains optimistic.

"Ultimately, the world is going to become one connected information society," Lefteroff says. "We still have to continue to build out the backbone and infrastructure for that."

But who is going to finance development of this next-generation infrastructure? The pressure is on larger companies. While industry innovators such as AT&T, Lucent and Telcordia Technologies don't publish how much they spend on research and development, they all say they are struggling to maintain their research budgets in the current economy.

Jeff Jaffe, vice president of research for Lucent's Bell Labs, says that for the last two years most network innovations came from start-ups. Now he predicts innovation will come from big companies that can build more reliable and scalable products.

"The crash and burn of the start-ups and the [competitive local exchange carriers] had some enduring benefits for the industry," Jaffe says. "It turned up the tempo of innovation, and that's something that we've very much responded to."

Jaffe says it is more cost-effective for companies like Lucent to fund their own research and development than to acquire start-ups. As an example, he points to an all-optical router that Bell Labs developed this year.

"The total investment in research for the lambda router was tens of millions of dollars," Jaffe says. "Compare that to the billion-dollar acquisitions of start-up companies in the telecommunications space, and you see that the internal model is much more efficient."

Lucent has been a victim of billion-dollar acquisitions gone awry, as the company recently shut down Chromatis optical transport operations, which it acquired last year for $4.5 billion.

Despite the efficiency of in-house research, many network companies are adjusting their research and development budgets to meet reduced revenue targets.

AT&T Labs is focusing its research dollars on a few strategic areas. This trend started a few years ago, "but it's much more significant now," Leighton says. "There's pretty tight alignment between the research program, the development program and the business units."

Leighton argues that a more focused R&D program benefits corporate network executives. "The problem with all the innovation going on in the network industry is that complexity was going up exponentially," he says. "Take IP VPNs, for example. We want to get to a single, simple offering that is easy for our salesforce to sell, easy for our customers to implement, and easy to manage."

AT&T Labs and IBM are among the network companies putting more emphasis on patenting innovations that come out of their labs. Indeed, patent applications in telecommunications have continued to rise in 2001, although at a slower pace than in the past (see chart, left).

Maintaining staffing levels is another challenge for network industry labs. Bell Labs currently employs 16,000 people, down from 27,000 a year ago, although many of them went to Lucent spinoffs. Telcordia Technologies has conducted layoffs because fewer service providers are enlisting the company's research expertise.

"[Service providers] have a pretty low appetite right now for paying for research," says Richard Wolff, vice president for advanced network systems research at Telcordia. "It varies from technology to technology, but the overall trend is downward."

Telcordia has shifted its internal research funds to focus on the most promising areas, such as wireless and e-commerce. Telcordia also plans to pursue more government research programs.

"Government funding is remarkably stable," Wolff says.

Many observers predict the federal government will play a more important role in funding network research than it did during the dot-com boom.

Federal spending on advanced network research is holding steady. Federal Sources, a market research firm, estimates that federal agencies will spend $1.31 billion on high-performance computing and communications research in 2002, up slightly from $1.28 billion this year.

The Clinton administration's Next Generation Internet initiative has ended, and President Bush's IT advisory council is reviewing all federal, networking research programs.

Patent graphic

For now, the most forward-looking projects are coming out of the National Science Foundation (NSF). The NSF awarded a $53 million grant in August to create the world's most powerful distributed computing facility. Developed with Qwest Communications, the facility will have a dedicated optical network that operates at up to 80G bit/sec. And an initiative to be announced this month will create a national test bed for evaluating the efficiency of next-generation middleware.

"All across science, we are seeing bigger and bigger data sets," says George Strawn, director of NSF's Computer and Information Science and Engineering Activity. "If you've got big data sets and millions of workstations around the world that want that data, that sounds like a requirement for an advanced networking infrastructure."

Research collaborations between government, industry and academia appear to be holding up. California's CalIT2 project, which is building test beds for emerging optical and wireless technologies, launched a year ago with $140 million in industry pledges over four years to match a $100 million state grant.

"In spite of a meltdown in the stock market and huge debt loads among wireless carriers in Europe, we have really seen no major backing away by industry from its support for our institute," Smarr says. Among the industry backers are Akamai, Compaq, Ericsson Wireless, IBM, Qualcomm and Sun.

Some observers say the slowdown in network innovation is caused more by the lack of user demand for advanced networks than by reduced research and development funding.

"How come the high-speed Internet hasn't burst on the scene? Because we need the same 10-year incubation for the high-speed Internet as we had for the low-speed Internet," Strawn says.

Strawn predicts it will be another five years before corporations develop applications that take advantage of the high-speed networks available today. "There are technical innovations coming out of the research labs ready to be put into use, but you've got to have the demand for them," he says.

Net innovation gets squeezed

Related Links