ORLANDO, Fla. -- NFOEC 2003 -- Former Cisco executive and InterWest Partners venture capitalist John Adler told the audience here that the new model for startups is to develop incremental improvements to a carrier's network – not a product that changes the way they do things.

"Disruptive equals death," said Adler in his keynote address at the 2003 National Fiber Optic Engineers Conference Tuesday.

Ironically, Adler was once was the senior director of marketing at Monterey Networks, which made a wavelength router so disruptive it was bought by Cisco Systems Inc. (Nasdaq: CSCO). Cisco reveled in Monterey's disruptiveness – until it realized how expensive it can be to try to sell a disruptive idea to conservative carriers, whereupon it discontinued the product (see Cisco Kills Monterey Router).

Adler said innovative technology can still make its way from startups to carriers, given the sour economy. "What has changed is how innovation is moving in the value chain," he announced. "But the fundamentals haven't changed."

Translation: The stuff of good startups is more or less the same. They need strong management. They need really slick technology. They need enough cash to get from PowerPoint slides to profitability.

But nowadays, a startup needs a carrier customer or a larger incumbent vendor to vouch for its technology early in its development. Said Adler: "Early proof of concept is essential now. You can't spend $30 million in 30 months looking for a customer."

Nevertheless, Adler remains optimistic that there are ways for telecom startups to make profits while still creating the customary 1,000 percent return for their investors. Naturally, this requires the startup to be a lean operation.

It also helps if the startup begins its life by providing products that are "modular upgrades" to elements of the carrier network – stuff that the giant incumbent vendors can't or won't develop because the revenue opportunity is too low. Such technology – bit it equipment blades and network software – can make it easier for the startup to find business partners and get acceptance into the carrier network, Adler said.

From there, it's up to the startup to carefully manage how it owns the customer relationship. After the carrier has bought into the incremental stuff, then the startup will stand a better chance of selling something that's a little more advanced and a little more aggressive.

Another approach Adler suggested is for startups to first sell to enterprises, and then move into the carrier world. Cisco wrote the book on that approach, he added, as did Newbridge Networks. And try targeting the access network, too: "Networking is all about moving bottlenecks around, and the access network is where the bottlenecks are now."

Adler's most recent bet as a VC is on Covaro Networks Inc., a company devising a way to extend Ethernet into the enterprise while making it more manageable as a service (see Covaro 'Jacks' Ethernet). It's hard to say if Covaro follows all the rules for successful startups that he discussed yesterday. It is safe to say one thing, though: It ain't no Monterey Networks.

— Phil Harvey, Senior Editor, Light Reading