Is Procket's science strong enough?

Since its founding three years ago, Procket Networks has operated in stealth mode, quietly creating advanced networking gear and collecting hundreds of millions of dollars from venture capitalists. This week the company started to let details of its operation trickle out, including plans to launch its first products later this year.

The release is expected to be closely watched by investors, competitors and executives at similar start-ups who are eager for a sign that telecom companies are once again willing to purchase pricey new equipment.

"They're growing up, morphing from a start-up to a little bit larger organization that will start producing products," said Infonetics Research analyst Michael Howard. "After all this hype they've built, the real test for Procket is to get in carrier networks.

"There's a certain amount of cynicism of, how do you enter the market against Cisco and Juniper and be different and attractive to carriers?" he added. "But they have the team with the potential to do that."

Perhaps the most valuable player on that team is Tony Li, a Procket co-founder who helped develop the software that runs on all Cisco routers. He jumped ship to Juniper in 1996, where he designed the software for its high-end router. Partly as a result of his work, Juniper has achieved a 30 percent share of the $2 billion market. Li this week was named as a director of Procket, along with Network Appliance CEO Dan Warmenhoven.

In addition, Procket's ranks include a former Sun Microsystems engineer who helped design versions of Sun's flagship UltraSparc processors. The company's CEO is Randall Kruep, who was lured from another networking company, Redback Networks, just over a year ago.

Procket's deep pockets
This brain trust is backed by an enviable war chest: Procket said Wednesday that it has raised $272 million in venture capital, a staggering sum considering that VCs have all but gone into hibernation as they await a thawing of the IPO market.

Based on that investment, Procket was valued at $1.55 billion last year, according to research firm Venture Economics. Although that valuation may better reflect the go-go days of the late 1990s than the harsh reality of 2002, Venture Economics said it ranked Procket as the country's highest valued VC-backed company.

The release of Procket's first products could be timely: A report last week by analyst firm Sanford C. Bernstein speculated that networking equipment sales could rebound by the second half of this year. The report's author, Paul Sagawa, accurately predicted the start of the current spending slowdown that has devastated networking companies.

Still, Procket has been dealt some short-term setbacks and its ultimate destiny may be determined more by macro-economics than the strength of its science. Co-founder Sharad Mehrotra recently left, and Bob Howard-Anderson, Procket's vice president of product operations, resigned this month to join another start-up, Occam Networks. Procket has also laid off some employees in recent months.

Meanwhile, there's no guarantee that the telecom industry--battered by overbuilt networks, massive debts and frugal consumers--wants or needs Procket's products. Indeed, some analysts are predicting further consolidation or outright failures among the company's prospective customers.

The telecom bust "makes it difficult for start-ups who must compete on having better technical ideas," said Dataquest analyst Tim Smith. "Service providers are dramatically cutting back their expenditures. The focus is increasing or maximizing cash flow, rather than moving their networks forward in any way."

Procket CEO Kruep acknowledged that "service providers have not made profits...But we intend to help them with profitability and deliver to customers an experience that is rich with speed, and consistent and predictable. We don't have that today."

He added that Procket this year will launch its first products, which he described as a mix of hardware, software and network processor technology that will allow telecom companies to offer new Internet services at a lower cost, as well as improve the performance of their networks.

Winnowing the field
If Procket does succeed, it will be against some tall odds. Howard, the Infonetics analyst, believes the networking market will consolidate before many start-ups have a chance to even launch their products.

"There will be fewer companies that survive," Howard said. "There's too many companies. There always is, but many companies are getting weeded out in the financing round, rather than get weeded out in the marketplace."

For example, IronBridge Networks filed for bankruptcy last year after failing to secure an additional $100 million in financing, one of numerous start-ups to fold after the market soured.

On the customer front, telecommunications service provider Metromedia Fiber Network--which Procket said was one of seven companies testing its technology and which was the only one it publicized--warned this month that it may have to file for bankruptcy if it cannot restructure its debts.

But Kruep said the company still has its original investors and has not had to devalue itself to attract funding, an issue other networking start-ups such as Caspian Networks has faced.

Others agree that Procket at least has a fighting chance.

"They have a fantastic team and the funding and are going into a space that can be highly lucrative. It's a bit dormant now, but all they have to do is listen to customers and execute," said a former Procket senior executive who declined to be named. "Procket is stable, and has the talent and money. They're coming out with very leading-edge technology. They're pushing the envelope in several different technology areas."

Part of the difficulty in handicapping Procket's odds is that little is known about those products. Some sources said the company is building next-generation high-speed routers, which telecommunications service providers could use to handle Internet traffic. Other sources noted the company could be focusing on one element of a routing system, such as a new chip or a new set of software.

Kruep would only say that Procket's technology is all of the above, and aimed at not only service providers but also large corporations with vast internal networks.

"We have our whole company focused on customer profitability and the ability to deliver reliable services to their customers. It's a much different approach than just the next hot box," he said, referring to routers.

Some setbacks
While remaining optimistic, Kruep acknowledged that the company has cut some workers, although it is hiring in other areas. Procket's Web site shows openings for 25 jobs, including chip and software designers.

There also has been speculation that Procket has run into trouble with its technology, but Kruep maintained that its plans are on track. Service providers have been testing out Procket's equipment for the past few quarters and will continue to do so this year, he said.

"The customers we've engaged have a good idea of our timing. Our timing is going to be announced over the course of the year," he said. "In this post-Internet world, customers are really focused on hearing people doing something, not talking about doing something."

Added Stuart Phillips, a general partner at U.S. Venture Partners: "Everyone loves to speculate, but Procket has maintained its focus on what it set out to do."

"There's no rush to tell people what they're doing. When they're ready, they'll tell people. They're still in stealth mode, and I'm pleased with the progress they're making," he said. U.S. Venture Partners is a Procket investor.

Despite the slowdown in telecommunications equipment spending, Phillips believes the future is still bright for Procket and other networking companies.

"Everyone says the sky is falling because capital expenditures are down, but $65 billion will be spent this year, and that's a fair bit of gear," said Phillips, who was Li's boss at Cisco.

Phillips believes that telecommunications carriers will start spending more money to build up their networks later this year or early next year. "For the last year, they've focused on the capacity they've had in place, rather than expanding," he said. "But pretty much all the network operators I talk to say, 'Yup, we're going to have to start building out and adding capacity again.'"

Smith, the Dataquest analyst, is less optimistic. He believes that 2002 will be a tough year and that growth in the network equipment market won't return until mid-2003.

When the market recovers, Phillips believes Procket--and other networking start-ups--will have a chance to succeed, much like Juniper in its early days, as long as the companies offer technological advantages.

"All the operators I talk to are willing to look at new companies if you can give a clear advantage over the equipment currently being used," Phillips said. "The value proposition that Procket set out to deliver is still every bit as valued as from the beginning."

But whether Procket's science can rise above the wreckage of the telecom sector is yet to be seen.