SAN JOSE, Calif. — One casualty of the wreckage surrounding the burst comms bubble will be good technical ideas, including some related to Internet Protocol, that never see the light of day.
That was one of several glum points discussed by analysts and venture capitalists in a Communications Design Conference panel session today (Sept. 25). Signs of hope persist amid the rubble, though, for companies positioned correctly. End-to-end portfolios are de rigueur, as the market will have little mercy for single-product companies, panelists said.
Regarding the lost generation of networking technology, panelists noted that champions of new Internet Protocol ideas — carriers such as WorldCom or the upstart competitive local-exchange carriers (CLECs) — can no longer pursue new technology due to a lack of money, resources or Wall Street permission.
"We have missed out on a two-to-three-year round of iteration and experimentation. A lot of those ideas will never be deployed, and we may never understand if they worked or not," said Greg Rossman, general manager of Pequot Ventures Inc.
A case in point is the terabit router, a glamorous high-capacity system championed by companies such as the now-defunct Pluris Inc. "Is the terabit router the right way to do the Internet? Nobody can tell you," Rossman said.
"That is one of the tragedies we're seeing in the IP [Internet Protocol] world," said Jeremy Bunting, analyst with Thomas Weisel Partners. "The technology never reached a maturity point where the CLECs could deploy it. So the CLECs never made revenues, and the CLECs failed."
Slim hope for narrow players
What's left is an environment of large customers who are skittish about buying from startups. For that reason, all panelists agreed that startups have to pursue end-to-end product portfolios if they're to have a substantial future.
"Point solutions are tough," Rossman said. "If you're going to have one, it's got to be incredibly high-margin and [address] an incredibly painful point."
Given the economic climate, such companies aren't candidates for initial public offerings, he said. "Those companies tend to be more like product lines than companies, and product lines tend to be acquired," Rossman said, noting that Pequot prefers to invest in startups that have a reasonable chance of fielding an IPO.
And acquisitions are scarce, because potential acquirers don't want to take the risk, Bunting said.
"If you've got a product that is in the hands of tier one [customers] — and it has to be tier ones signed up for six months of revenue — I think a deal could possibly be done," he said. "If you don't have that, the company just burns through the cash it has and it dies, sadly."
David Silverman, a partner with venture firm 3i, added that existing sales and revenues don't necessarily help a company get bought. "It's a buyer's market, not a seller's market, so it doesn't really matter what your statistics are," he said.
One audience member asked if a venture capital firm could combine its portfolio companies to create an end-to-end play, but the panelists frowned at the idea.
"You want to be careful about combining early-stage companies because the laws of probability are against you. If one has 0.8 [80 percent chance to succeed] and other one is 0.8, you've got 0.64 right there," said Dushyant Desai, an analyst with U.S. Bancorp Piper Jaffray.
An even worse idea is to combine a startup with one of its potential suppliers, to create a super-vertical company. "That is only self-serving," Desai said.
Overall, it could be a good time to launch a startup, given that investors' expectations will be more realistic than during the bubble. In particular, they won't be looking for a company to get acquired quickly, Desai said.
Of course, panelists didn't exactly recommend that audience members quit their jobs. "It's probably also not a bad time to think about staying where you are right now, because we don't think the semiconductor industry in the next 12 to 18 months is going to experience any great boom," Rossman said.
