Will Wi-Fi Fly?

Monday April 18, 2003

Wireless networks are proliferating, but making money will be tricky

IN SILICON VALLEY, there’s a desperate search for a dramatic new… something. Something that will drive chip, hardware and software sales. Something that will create jobs and fill empty office buildings. Something that will pry open the initial public offering window. Something that will shake the tech sector from its three year funk. Now, there’s a spreading belief that the something just might be Wi Fi.

Short for “wireless fidelity,” Wi Fi certainly is a red hot technology. It’s also an appealingly simple idea: fast Internet access without wires. Also known as 802.11, for the technical specs set out by the IEEE (the Institute of Electrical and Electronics Engineers), this radio technology really is going to change the way many people connect to the ‘Net.

It’s already happening, in fact. Consumers are installing wireless networks at home, so they can Web surf from the couch, the porch or the backyard. Companies are installing wireless networks, too, freeing employees from their desks. And you can now read e mail while sipping a latté at one of thousands of public “hot spots” — the wireless Internet nodes found in coffee shops, burger joints, hotels, airports, bookstores, copy shops and convention centers.

Look further out, and the potential grows. Wireless networks could eventually provide stiff new competition to DSL lines and cable modems for local access to the Internet — technologies under development would dramatically increase Wi Fi’s range beyond its current few hundred feet. New techniques now in the lab would allow seamless handoffs between nodes, the way mobile phones now switch between cells. Combine Wi Fi with Internet based telephony, and you have an intriguing potential competitor for existing cellular networks.

While there are still real hurdles to widespread adoption, key players in the tech food chain nonetheless have been making big bets on wireless networks. Intel, for instance, has committed $300 million to push Centrino, its brand name for a package of chips, including a microprocessor and Wi Fi components. For the semiconductor giant, it’s the largest branding campaign since “Intel Inside.” The chipmaker has also pledged to invest $150 million in Internet start ups: It’s already backing more than a dozen Wi Fi related newcomers. By the end of 2004, Intel contends, most new laptops will be Wi Fi ready.

Cisco Systems, meanwhile, recently inked a $500 million stock deal to buy Linksys, a leading producer of consumer Wi Fi hardware. Microsoft has a stake in the market as well: Seeing a way to rev up PC sales, the company has jumped into the home networking hardware market in competition with Linksys. Meanwhile, AT&T, Intel and IBM are all backing Cometa Networks, a San Francisco start up that plans to offer wholesale wireless Internet access via established telecoms.

While that’s all encouraging, Wi Fi poses a conundrum for equity market investors. While Intel, Cisco and Microsoft would all benefit from the widespread adoption of wireless networks, none of these companies exactly qualifies as a pure play. And there’s a more fundamental problem: While sales of wireless devices are expected to soar in the next few years (see charts below), the segment faces growing competition, falling prices and shrinking margins — along with spreading hype. In a special supplement to its May edition dubbed “UnWired,” Wired magazine gushes that Wi Fi is a “street level uprising with the power to change everything.” Sound familiar?

In another eerie parallel to the Internet bubble days, high hopes for Wi Fi have set off a feeding frenzy in the venture world. Venture firms have funded several dozen Wi Fi chip start ups, which find themselves going head to head with more established players like Intersil, Agere, Texas Instruments and Broadcom, and with Intel’s Centrino debut. Low cost Taiwanese chipmakers are entering the market as well. Result: commoditization and steep price erosion. (It’s telling that none of Intel’s Wi Fi investments to date are in other chip companies.)

Sliding chip prices, in turn, have triggered big price cuts for both Wi Fi access nodes — the little boxes with an embedded radio that you attach to a DSL or cable connection — and for the cards that you slide into PCs to wirelessly connect to those nodes. You can now buy access points for under $100; Wi Fi cards often go for less than $50, and sometimes less than $20. And prices keep dropping. That’s certain to trigger consolidation: More than 170 companies are providing more than 450 Wi Fi certified products, according to the Yankee Group, a Boston based tech research firm. That’s far more than the market can support.

Those falling prices, however, are the key driver for Wi Fi’s growth. Combine cheap hardware with a technology using unlicensed (read: free) radio spectrum, and you’ve got a recipe for rapid adoption.

The technology itself isn’t really that new: 802.11 based products have been sold since 1997; the 802.11b standard that dominates the market was adopted in 1999. And there are other technologies for wireless access. For years, it has been possible to get slow speed connections via cellular networks; and the promise of so called “3G” cellular networks remains in improved data services. And perhaps you remember Metricom and its Ricochet service: The Paul Allen backed wireless Internet system, launched in 1995, burned through more than $500 million before going bankrupt in 2001. (Later acquired by privately held Aerie Networks, Ricochet service is still operating in Denver and San Diego.)

After a relatively slow start, however, Wi Fi seems to have reached the tipping point. “The great value of Wi Fi is not that it is fantastically advanced technology,” explains Kevin Werbach, a consultant and commentator who spent four years at the Federal Communications Commission, working on technology policy issues, and later edited the newsletter Release 1.0. “It has many limitations, and it’s not near the cutting edge. But it’s being widely adopted. You can buy a Wi Fi card for $50. And every time someone puts up a hot spot, I benefit from that. Every added bit of connectivity adds to the collective benefit.” That’s a variation on Metcalfe’s Law — the idea proffered by 3Com founder Robert Metcalfe that a network’s utility increases exponentially as the number of users grows.

That dynamic drives growth. In 2004, according to Oyster Bay, N.Y. based tech market research firm Allied Business Intelligence, more than half of all computer notebooks sold worldwide will be Wi Fi capable; by 2008, the total is expected to reach 93%. Although lagging a bit behind PCs, the same thing is happening with personal digital assistants, or PDAs — 50% of handhelds should be Wi Fi enabled by 2006, according to ABI, up from 3% in 2002. ABI figures all the Wi Fi enabled hardware will drive up the number of paying hot spot customers to 5.5 million in 2006, from an estimated 24,000 last year. Rich Beyer, CEO at Intersil, the leading producer of Wi Fi chips, contends there will be at least 100 million Wi Fi capable systems in the market by year end 2004.

“As the costs come down, it becomes virtually free to put Wi Fi radios in a wider and wider variety of devices,” Werbach says. “Pretty soon, you won’t be able to buy [the devices] without Wi Fi. Right now, $10 $20 for a Wi Fi chip doesn’t add much to the cost of a laptop, but it’s significant for a PDA. At $2, it doesn’t add much to a PDA’s price. At 50 cents, every consumer electronics device will have a Wi Fi chip.”

Unfortunately for chipmakers, the soaring unit sales are being offset by those plunging prices. InStat MDR expects Wi Fi chip revenues to peak in 2005, then shrink, even as unit volumes keep growing, as a result of sliding prices. Then who else will make money on this? Software and services providers are the most obvious beneficiaries, but not the only ones.

Wi Fi actually includes three distinct end markets — corporate networks, home networks and public hot spots. The corporate market is by far the largest, accounting for about three quarters of the $1.88 billion spent on Wi Fi equipment in 2002, according to tech researcher International Data Corp. of Framingham, Mass. It’s the slowest growing of the three markets, however: IDC sees a 31% gain, from $1.44 billion in 2002 to $1.89 billion in 2004. Keith Waryas, research manager for wireless at IDC, says adoption has been slowed by both shrinking corporate IT budgets and concerns about the vulnerabilities of wireless networks.

Waryas maintains that wireless network security isn’t really as big a problem as some fear, but concerns linger nonetheless. The most common fear is that some evildoer with a laptop will park outside corporate HQ and poke around the network, reading the CEO’s e mail, stealing customer data or worse.

WHILE INFORMATION TECHNOLOGY departments may be reluctant to add wireless networks, the technology is gradually being adopted anyway, driven from the bottom up by employees eager to untether. Indeed, the low cost and easy installation of wireless access points has spurred a proliferation of “rogue” nodes on corporate networks by employees determined to cut the cord. That practice is dangerous: Nodes set up inside the corporate firewall offer a $100 way to get around thousands of dollars in corporate network security measures.

The adoption of corporate wireless networks will be tied in part to the rise of another emerging technology: Internet telephony. Avaya, for instance, has begun pushing wireless business phones, which can make you reachable anywhere with Wi Fi access, whether on the corporate campus or from a public hot spot. Make it a dual mode phone, with the ability to use cellular networks as well, and you can have one phone and one number, inside the office or out. Telesym, another Intel backed startup, sells software that turns PDAs or laptops with Wi Fi access into Internet based phones.

“Telephony is just another Internet application,” says John Patrick, a consultant and author who until his retirement a year ago was IBM’s chief Internet guru. “The telecom industry thinks of the Internet as one of the things you can do with telephony services, but it is exactly the opposite. Telephony is one of the many things you can do with the Internet.”

Home networking, while smaller than the enterprise market, is growing rapidly. International Data Corp. expects hardware revenues to rise from $306 million in 2002 to $539 million in 2004. Growth is limited in part by the number of households that have both broadband access and multiple computers, but those numbers are growing. Cisco’s decision to buy Linksys, despite margins far lower than those available from the company’s enterprise routers, shows the networking giant’s enthusiasm for the potential of home networking.

The promise of wireless home networks is that they will do far more than link PCs to the ‘Net. Coming soon, say Wi Fi enthusiasts, will be a wide range of 802.11 enabled consumer electronics devices, zapping digital audio and video around your house. Televisions. Stereos. MP3 players. Set top boxes. TiVo, maker of set top personal video recorders, has announced a Wi Fi enabled model. Sanyo has demonstrated a Wi Fi capable digital camera.

Perhaps the most intriguing — and potentially most over hyped — segment of the market involves the public hot spots, which are spreading like wildfire. International Data Corp. sees equipment revenues for public access Wi Fi growing from $127 million last year to $502 million next year; IDC expects the number of commercial hot spots to jump to 9,700 this year, from 3,700 in 2002. By 2006, IDC figures, there should be close to 50,000 of them. And that doesn’t count free access sites, home networks or corporate nodes.

Yankee Group figures there will be 5.4 million regular users of paid Wi Fi networks in North America alone by 2007, with business users representing 86% of the total. It’s worth noting that Research in Motion has just slightly more than a half million users for its Blackberry pagers, which dominate the market for e mail capable handheld devices. Wi Fi service providers are betting that at least 10 times as many people will be willing to pay for more complete remote Internet access.

Still unsettled is just how much the service providers can charge. IDC’s Keith Waryas contends that Wi Fi should eventually be bundled with Internet access or other services, rather than being billed as a stand alone product. That’s not quite happening yet, however. All of the providers offer a range of plans. Boingo, for instance, offers service at $24.95 for 10 days of usage a month, $49.95 for unlimited access, or $7.95 for two days of use. Wayport charges $6.95 a day for an airport connection, $9.95 a day in hotels, and offers a variety of monthly and pre paid plans. T Mobile’s rates range from $29.95 a month for annual contracts to metered access at 10 cents a minute. On March 1, T Mobile chopped its monthly rate by $10 a month, and cut the cost for a single day of access as well. Some pundits theorize that retail locations with Wi Fi might ultimately cut their rates a lot more or simply provide free access to lure customers.

Indeed, free Wi Fi networks are popping up around the country; some retail outlets are ordering a DSL line, plugging in an access node and then providing free access. You can find that, for instance, at Buck’s, the legendary venture capitalists’ restaurant hangout in Woodside, Calif. Owner Jamis MacNiven hasn’t gone out of his way to advertise free Wi Fi access, but says there’s “always somebody wired in at one of our tables.” MacNiven says he incurs few costs beyond the $60 a month Internet connection that he uses himself anyway and since he’s not charging, there’s no billing to worry about. He thinks the free access attracts some extra customers. And, in any case, he contends that charging for access “would be like charging extra for salt and pepper.”

Don’t expect the free model to dominate, however. “Business travelers with 30 minutes to check e mail don’t want to go looking for war chalking,” says former IBMer John Patrick — meaning they don’t want to take time hunting down free access points. “I happily pay T Mobile for that reason. When I get to an airport, I head to the Admiral’s Club, log in, and as soon as I hit the browser and enter a password, I’m a happy camper.”

Pricing isn’t the only difference between the Wi Fi service providers; they also have a range of business models. Cometa has set itself up as a wholesaler, offering service through larger telecom service companies. Larry Brilliant, who until recently was CEO at Cometa, says the firm “will never have a brand.” The Wi Fi customer relationship should be owned by the existing carriers, “unless they screw it up,” Brilliant says. “For AT&T to add Wi Fi, there are no customer acquisition costs to speak of. They don’t have incremental billing issues.”

Cometa’s strategy in part reflects its belief that the hot spot market will be paid for largely by corporate users. Brilliant contends that information technology departments will willingly pony up for secure and reliable remote access for both laptop lugging, airport hopping road warriors, and “windshield warriors” salesmen, insurance adjusters and support people who do most of their travel behind the wheel. Cometa expects to have 5,000 hot spots by the end of the first quarter of 2004, with 20,000 more 24 months later. Among Cometa’s targets: Big Mac eaters. McDonald’s recently launched Wi Fi access from Cometa in 10 Manhattan outlets. The burger giant plans to have the service in 300 restaurants by the end of the year.

Boingo, started by EarthLink founder Sky Dayton, is set up as a network aggregator — it has pulled together access to more than 1,100 hot spots from a variety of smaller providers — and offers service both directly to consumers and via carriers and Internet service providers. iPass, a Redwood Shores, Calif., company that provides aggregated dial up access to corporate business travelers — Barron’s parent Dow Jones is a customer — has expanded its service to offer aggregated hot spot access. iPass, which has filed to go public, claims to have more than 1,000 hot spots. GRIC Communications, an iPass rival, says it has 1,200 combined wired and wireless Internet access points.

Increasing the availability of access points will be crucial to winning customers. As Kevin Werbach notes, “there’s not much value in connecting one airport. There’s value in connecting every airport.” Success on that score often has less to do with cost than with the business negotiations between the provider and facility owner.

In some cases, that requires creativity: T Mobile, for instance, has cut a deal to place nodes alongside the televisions showing CNN in airport terminals. T Mobile, a Deutsche Telekom subsidiary once known as Voicestream Wireless, has one of the largest existing networks more than 2,000 hot spots most famously in Starbucks coffee shops. It has been by far the most aggressive of the cellular networks in the Wi Fi market, a fact that reflects the ambivalence about the technology in the cellphone business. While Wi Fi poses a complication to plans for wireless Internet access using 3G cellular networks, the mobile operators,with their existing customer billing relationships, have an obvious advantage over upstarts.

“My sense is that every major cellular company has a group of people who see this as a major opportunity, and another group saying it’s a threat,” Werbach observes. “The first group eventually will win. They can’t roll over and pretend it’s not there.” It’s hard not to be skeptical about Wi Fi’s investment potential, given the scars that investors received when the tech bubble exploded. Cometa’s Brilliant, for one, considers the high hopes for Wi Fi a little unsettling. “It bothers me and frightens me when the entire American economy is looking to one very modest new technology to save it,” he says. “That’s not a very good thing. 802.11 is a modest technology; it’s wireless ethernet, that’s all it is. It’s not a panacea.”

A more troubling view comes from Brewster Kahle, founder of Alexa, the Internet search service now owned by Amazon.com, and the head man at the Internet Archive, a project to keep a complete record of the Web. Kahle, who since 1997 has been experimenting with free wireless Internet access in San Francisco’s Presidio district, thinks that a failure to develop widely available Internet access fast enough to handle full motion video could cripple the rest of the tech business. “We need Moore’s Law of Internet bandwidth,” he says, referring to the idea laid out by Intel co founder Gordon Moore that the number of transistors per chip doubles every year or two.

Consider, he suggests, disk drive capacity. You can get a 250 gigabyte hard drive for $300. If you were downloading 24 hours straight on a DSL line, you would only fill about 7 gigs, he notes. “Why would anyone want a 500 gig drive, or a 1,000 gig drive, if you can’t fill the 250? And if we lose the momentum of Moore’s Law, I’m not sure you get it back again. The tech sector could just tank. How much faster do you need to run Microsoft Word?” Kahle thinks that very high speed wireless access — fast enough to handle DVD quality video — could be the answer.

But John Patrick contends that Wi Fi really might be the next big something the Valley desperately needs. “We’re very much in parallel with where we were with the Internet almost 10 years ago,” he says. “I remember looking at the Internet at IBM in 1994 and thinking: ‘This is really cool, but where’s the money?”’ The questions people have about Wi Fi now are the same ones we had in ’93 and ’94 about the Internet. Skeptics say it doesn’t scale, it’s not secure, it’s not industrial strength. It’s the same things people said about the Internet. But there’s no stopping Wi Fi. It’s a grassroots technology, totally distributed, standards based, global, with nobody in charge. Those are the reasons the Internet has flourished. And the implications are huge.”