The treasure and terror of the current state of technology investing will spill into view this week when Intel ( INTC) formally introduces its Centrino chipset for mobile computing with a splashy television- and print-advertising campaign. The semiconductor giant has vowed to commit more money to promoting Centrino than it did to promote the Pentium 4 -- a magnitude of ad spending that suggests desperation to change its public perception from a company known for making faster and faster desktop computer products to a company known for making products that release us from the desktop. If this is truly an epochal shift in technology from the dead end of a wired past to the limitless horizons of an unwired future, you would think that investors would be salivating over the opportunity to make money in myriad ways. Yet, except for highly diversified large-caps like Intel and Cisco Systems ( CSCO), for which wireless networking continues to be just a promising side project, it's still hard to figure out where high-margin new business will get done. Free Wi-Fi for Everyone The problem is that wireless local-area networking -- commonly known today by the cute marketing term "wi-fi," short for wireless fidelity -- increasingly looks like the most destructive new technology to come along in recent memory. It is so inexpensive to deploy as a means to access the Internet at high speeds, and use is so addictive, that its very popularity could be its downfall. The paradox is rooted in the weak overall economy. For rather than deploying wi-fi as a means to make money, several large companies recently have announced plans to give it away to build loyalty for their faltering core services. And once this genie is out of the bottle, it's hard to imagine how anyone manages to stuff it back in.
The leader in this trend is Omni Hotels, a privately owned chain of 40 luxury hotels based in Texas. But other hotel chains, such as Wyndham Resorts, have followed suit, as have the McDonald's ( MCD) and Barnes and Noble ( BKS) retail chains. In February, Omni announced it had installed high-speed wireless networks at its Los Angeles, Dallas, Chicago and New York locations and planned to install it at other locations by year-end. If you don't have a notebook computer that's already equipped either with a wireless network interface card or a wireless chipset like Centrino, the hotel will lend you a wireless adapter that attaches to your Ethernet connection. When you can pick up a piece of cutting-edge technology for the same price as a complimentary toothbrush, you have to figure that something is desperately wrong with the business model. Free wi-fi in hotels is a remarkable first step on a path that will probably progress at warp speed, much like cable TV in the 1970s. For a few years, hotels and motels advertised cable to visitors because it was a differentiating luxury, and they charged a pretty penny. Now it is a standard feature of rooms. From High Margin to No Margin How did wi-fi shoot down the path of high margin to no margin in record time? John Yunker, telecommunications analyst at Pyramid Research, said it happened because wi-fi is built on a standard that gives every hardware and software maker an equal opportunity to make the compliant products. You can't deviate from that standard and still have your products work in the network, so there's little opportunity to innovate cool differentiators. Since wi-fi is one of the few growth areas for technology makers worldwide, virtually every major and most minor chip companies have crowded into the field. The competition has driven down prices for a notebook computer wi-fi card from $200-plus a year ago to around $50 today, en route to less than $30.
High-end chipmakers Broadcom ( BRCM) and Intersil ( ISIL) are shipping wi-fi chipsets today based on a higher-speed standard than Intel's -- 802.11a and 802.11g, to get technical -- but the effort isn't expected to do them much good because the extra oomph is practically indiscernible in today's applications. Despite becoming the high-end wi-fi leader, Broadcom shares have tumbled 68% in the past year. And despite being first and holding several key wi-fi patents, Intersil shares have fallen 53% in the past year. Unlike Intel, neither is profitable. Eric Gomberg, a chip analyst at Thomas Weisel Partners, points out that investors now actually penalize Intersil for having a wi-fi division. "I don't think that's fair, but that's how Wall Street is treating the stock," he said. "In a bull market, having a dominant position in a hot technology that could be a killer app is a huge plus. Today, the Street says price pressure will only worsen from here, making wireless chips a profitless growth engine." Gomberg has gone so far as to recommend Intersil spin off its WLAN business so that investors can focus on its much more profitable analog chip business. If there's no money to be made in the chips that go into the consumer products, how about the big wireless transmitters, amplifiers, routers and antennas? Sorry, no luck there, either. In a just-published research report on wi-fi in the hospitality industry, Yunker said most hotels are buying Cisco products for their reliability and relatively strong software-based security. The equipment's not expensive, so why bother messing with a lesser brand? A hotel can install wi-fi in 175 rooms for around $30,000, about the cost of a single full-page ad in a metro newspaper. The just-buy-Cisco movement -- at hotels and elsewhere -- has pushed out virtually all other vendors. Agere Systems ( AGR.A), a former powerhouse in the business, sold its wireless LAN business to Proxim ( PROX), which announced last month that it had lost big hunks of market share to Cisco. Proxim has responded to its loss by cutting prices by a third -- never a good strategy for long-term survival. Proxim, mentioned by every analyst as one of Cisco's few competitors in the enterprise space, has seen its shares sink 83% in the past year to 46 cents. (Bargain hunters beware: It is swiftly running out of cash.)
No Opportunity on the Last Mile? The other place that wi-fi is getting ready to grow with no discernable benefit to investors is in the last mile, or local loop. Today, if you want broadband access at home, you have to pray that either your regional Bell operating company, such as Qwest Communications ( Q) or Verizon Communications ( VZ), or cable company has deigned to string a specialized line to your neighborhood from a nearby central office. This typically means that if you live anywhere outside the dense urban or suburban core, you're out of luck. And even if you're among the lucky ones, prepare to pay upward of $40 a month for sometimes-spotty service at speeds of around 256k to 768k. In an attempt to blowup this model, several well-financed start-up companies have already begun to sell advanced wi-fi transmitters, amplifiers and antennas that allow just about anyone to become a freelance telecom carrier by broadcasting a single broadband signal across a wide area. The pioneer in this field is Glenn Wilson, president and founder of M33 Access in northern Michigan. Frustrated that none of the telecom carriers would bring high-speed Internet connections to the depressed countryside surrounding his hometown of Rose City, Wilson figured out how to buy a hodgepodge of parts from Florida-based online retailer HyperLink Technologies and rig up the country's first renegade high-speed wireless ISP. He now operates the largest consecutive wireless broadband grid in the country -- about 100 square miles -- with no telecom carrier involved, by beaming the signal from one self-rigged 200-foot-tall tower to the next. He sells wireless broadband to farmers, homes, school districts and small-business owners for as little as $25 a month (if you have a child in school, he'll cut the price to $15). In a few weeks, Wilson plans to be able to add voice-over-Internet to his list of services, which will allow him to offer virtually free local and long-distance service at 4 cents a minute. "Our company is about bringing Michigan out of Hooterville," he says, referring to the '60s-era TV show "Green Acres" in which communications was handled by tin can. "We're causing the regional Bells some problems. We have 6,000 customers and they love us -- no one is going to move in and take them away."
Wilson uses Cisco routers at his head end and Orinoco-brand wi-fi radios made by Proxim; he runs the whole thing on Intel-based servers. But the whole operation, which obliterates the need for million-dollar switches made by old-school manufacturers like Lucent Technologies ( LU) and Ciena ( LU), isn't going to make equipment investors rich. The stuff is just too cheap, not to mention easy to upgrade and maintain. Jamie Cusano, sales manager at Hyperlink Technologies, said he has Federal Communications Commission-certified kits that will allow anyone to broadcast a wi-fi signal as far as 20 miles. "You can buy $1,000 worth of equipment and circumvent hundreds of thousands of dollars in local-loop costs," he said. "This has bigger potential than most people realize." With prices like that, it's easy to see how early efforts to charge a premium for high-speed public wi-fi "hot spots," such as the T-Mobile-powered spaces in most Starbucks ( SBUX) coffee shops, are likely doomed to the scrap heap of digital history. What's to prevent a mall operator from buying a high-powered antenna from Hyperlink -- perhaps one of the new ones coming from an interesting San Francisco start-up called Vivato -- and blasting wi-fi throughout its property as a loss-leading enticement for shoppers? An offer such as "Shop at the SuperMegaMall and Get Free Internet Access" would annihilate the Starbucks wi-fi model, where access currently costs a whopping $30 per month. If it's cheap and easy to make virtually every major retail location "hot" to attract diners, shoppers, readers, train riders and moviegoers, then there is no such thing as a hot spot. No spot, no geographically based service-business model as presently conceived. Keys to Killer Apps: Security and Convenience
The killer apps for investors will probably be software that enhances security and permits seamless roaming between all types of wired and unwired networks. One company with a shot at being successful is iPass, an enterprise-focused company that aggregates a worldwide set of Ethernet, 2.5G, wi-fi and dial-up networks into a single service. iPass sells its service to companies that want traveling salesmen or executives to be able to sign into the Internet securely through one simple interface on their mobile computer, no matter whether they're in Singapore or Cincinnati. iPass just signed a major deal with Cometa Networks -- the nascent nationwide wi-fi service backed by Intel, IBM ( IBM) and AT&T ( T) -- to share resources, and plans to make an initial public offering of its shares later this year. On Monday, Intel announced investments in four other wi-fi startups, including Vivato, that could be IPO candidates in the next couple of years.
So, does the cheap, plentiful equipment doom the likes of Cometa Networks and earlier hot-spot aggregators, such as Wayport and Boingo Networks, from ever making a decent buck? Not according to David Isenberg, an independent telecommunications analyst whose views can be read at Isen.com. He said in an interview last week that he believes a wide variety of wi-fi business models will develop in ways we cannot imagine today -- and that we'll ultimately become accustomed to paying different amounts based on variables such as speed and convenience. Consider, he suggested, the different ways we've become accustomed to paying for parking a car: Sometimes you put a coin in the meter, sometimes you rent space by the month, sometimes the merchant provides a free space outside its property, sometimes the merchant validates at a remote lot because it has no space of its own, sometimes you hunt for a free spot on the curb, sometimes you're willing to pay a lot to park close to a sports event, and sometimes you park illegally and eat the ticket. "Whether we end up having free wi-fi at home and in malls, paying a little extra for a higher speed at Starbucks, or stealing someone else's connection at other times, many different business plans make sense," he says. "There's nothing that says you need a single model for any given technology. I'm sitting on the edge of my seat waiting to see what's going to happen next; it's one of the most exciting things happening today in communications." Investors are on the edge of their seats, too, but the fabric is getting thin. For now, it looks like your best bets are Intel and Cisco, so long as they attract you for more lucrative reasons as well.