Wednesday, December 06, 2006

CONTOS AMERICANOS : THE NEW DOT-ECONOMY

SAN MATEO, Calif. -- Allan Leinwand's start-up, Vyatta, sells networking products for one-third the price of those sold by his former employer, Cisco Systems.
In the same office park, serial entrepreneur Munjal Shah recently launched an Internet search engine for images, a gamble that would have been prohibitively pricey just a few years ago.

Half a mile away, tiny Zimbra is going head to head with mighty
Microsoft by peddling cheap e-mail programs to businesses.
Across Silicon Valley and beyond, young companies are capitalizing on a freefall in technology costs to enter markets considered too expensive to brave as recently as two years ago. The resulting boom in new Web sites has been widely chronicled, but the economics making it possible have stayed mostly out of sight.
Computer costs have been falling inexorably for decades, driven down by steady improvements in the tiny chips that serve as their brains. But new factors are compounding those gains, unleashing a second wave of Internet start-ups.
Computer chips are being paired in new ways to dramatically expand their abilities. At the same time, the free-software movement born in the 1990s is arming a new generation of entrepreneurs with powerful development tools -- at little or no cost. Other advances have slashed the cost of high-speed Internet access, another key expense for start-ups. Today, a Web site pays one-twentieth of what it did six years ago to reach visitors over its broadband connection.
These cascading cost reductions are transforming every layer of the computer industry, from components inside machines to the arcane innards of the Internet and all the way up to e-mail software and Web applications used by millions of people.
"The threshold for starting a young Internet business has never been lower," said Ram Shriram, a venture capitalist and an original board member of Google.
Shah, who launched the online shopping service Like.com last month, said his company could not have existed five years ago because it would have cost too much to develop the technology for scanning and analyzing the millions of product images on the Web.
"There are structural reasons this boom is happening," he said.
His firm, Riya, paid about $50,000 for its software by relying mostly on free, open-source programs for its operating system, Web server and database manager. That compares with the $4 million Shah spent on software seven years ago when he started his first online company, which made software for managing eBay auctions.
Shah said the cost of the servers powering his new Web service is one-tenth of what it would have been a few years ago, a reduction made possible by the new ways computer chips are harnessed to work together. And instead of paying for advertising, Shah looks to the free buzz of the Internet to attract customers.

Like.com is exploring a new idea -- using visual recognition software to automatically find images on the Web that resemble each other. Other start-ups are pursuing older ideas using cheaper technology, exposing such industry giants as Microsoft, Cisco, SAP and Oracle to unprecedented competition.
In the early years of the Internet boom, established companies came to fear the rise of rivals with new ideas, at times just a couple of young engineers working out of a dorm room or garage. Now, a new idea is no longer essential, because the massive cost reductions mean start-ups can enter the marketplace simply by offering cheaper wares.

"When factors of production are in flux, they require new mind-sets," said Robert Metcalfe, general manager of Polaris Venture Partners and an inventor of the Ethernet networking technology. "That's an opportunity for new companies who get it. It doesn't necessarily mean they win and that big companies squander the opportunity. But it happens."
The three friends who founded Zimbra in 2003 during brainstorming sessions in a California coffeehouse saw the potential of using free software tools. They used this software, developed by volunteers and available on the Web, to create Zimbra's e-mail product and began selling it in February for barely two-thirds of what Microsoft charges. Since then, Zimbra's software has reached more than 4 million paid e-mail accounts at more than 1,000 businesses and organizations, including universities and
H&R Block.
Zimbra has raised about $30.5 million in venture capital, including $14.5 million earlier this year. But its costs have remained so low that it has not spent any of the $14.5 million -- and still has money left over from previous investment rounds, said chief executive Satish Dharmaraj.
Microsoft remains the heavyweight of business e-mail, with 140 million users. But some analysts think the shifting economics eventually could pose a problem for Microsoft and other industry leaders. For one thing, incumbents with profitable businesses face obstacles to exploiting lower-cost technologies. Most large software makers, for example, sell programs based on proprietary code. Exploiting free, open-source software to sell cheaper versions could cannibalize their established products and cut profits, provoking resistance from shareholders and employees.
"They're going down a road that has big, deep ruts in it. Getting out of those ruts is very hard," said Forest Baskett, a general partner of New Enterprise Associates, a venture capital firm.
Yet the giants of technology are hardly standing still. Some analysts point to Cisco, in particular, as a company trying to adapt.
For years, Cisco has called itself "the plumber of the Internet" because it makes most of the technology that routes data through local and global networks. Its high-performance devices called routers dominate their industry and bring in nearly $6 billion in revenue annually. Cisco and its rivals, including
Juniper Networks, mostly sell router packages that combine hardware and software into a single product.
But Vyatta, founded two years ago by Leinwand, slashed the cost of routers by selling the software and hardware separately. It exploited computer-chip advances that have allowed cheaper machines, such as those sold by Dell, to support routing, and lowered software costs by using free development tools.
As a result, Vyatta's customers can pay $1,000 for a Dell computer and another $500 for an annual subscription to Vyatta software -- a sharp discount off the prevailing router price of about $4,500.
While Cisco's routers still dominate the market, the company has not ignored the growing demand for cheaper products.
Three years ago, as more people set up computer networks at home, it bought a 15-year-old company called Linksys that was a leader in low-cost routers and turned it into a division that sells the devices using open-source software and off-the-shelf processors -- not unlike Vyatta. Mike Volpi, a senior vice president at Cisco, estimated that Linksys now commands about half of its market segment.
"We have been challenged by many start-ups in the past and managed to get through," he said, playing down Vyatta's attack on the market.
Leinwand set up Vyatta two years ago after colleagues at other companies told him they were looking for cheap routing software to run on standard computers.
By the time Vyatta got its first customer, in July, its workforce was still fewer than 20 people. The marketing chief took the first order on a Post-it note.
Among the initial customers was Adify, an online advertising service that tried a Vyatta router in a low-risk part of its operation in July, even though its executives were skeptical of open-source software. The trial went so well that Charles Stewart, Adify's vice president for operations, said he ripped out a strategically located Cisco router in November and replaced it with Vyatta's product.
Kelly Herrell, Vyatta's chief executive, acknowledged that Cisco and Juniper still hold an advantage at the top end of the market, where standard computers cannot yet support the most sophisticated networking needs. But in the middle of the market, standard hardware and Vyatta's software can satisfy the needs of most businesses, he added.
"I'd rather start at the bottom and work my way up," he said.

No comments: