Wireless: Great Tech, Prohibitive Cost
Or maybe not.
Bernstein Research analysts have peeked behind the wireless screen to determine whether the media and technology worlds eager for growth are expecting too much from a wireless "third pipe." Wireless data is the single most important growth engine for domestic telcos--and the this will be the "critical enablers of a vast range of new and planned technology products and services, from mobile video to location-based search," according to analyst Craig Moffett. Fueled by low-cost computing devices such as netbooks and smartphones, as well as the incoming Obama administration's aggressive high-tech policy, there is a growing expectation that wireless will emerge as an alternative to the wired broadband duopoly, the vaunted "third pipe."
Moffett's deep dive into the economics of providing truly high-capacity broadband services domestically over wireless networks has a bright future, but would be prohibitively expensive. The roughly estimated cost would be more than $200 billion in equipment, site acquisition, network engineering, installation labor and other non-equipment costs over the next five years. Such an investment would be more than double the amount of aggregate wireless capital expenditures made by Verizon Wireless, AT&T Mobility, and Sprint Nextel over the past five years.
"A network ten times the size of today's network would be economically unthinkable," Moffett said, requiring at least two to three times more than current revenues to support a ubiquitous data-centric network. The fewer wireless broadband users, the more expensive the service becomes for subscribers.
Advanced wireless data service beyond basic texting is at about 9% penetration today, and is projected to grow to 30% by 2012. The average bandwidth usage for each of those subscribers is less than 100Mbs per month (representing mostly BlackBerry and other email devices), or just a fraction of what a user could do with a wired connection. Moffett says that an iPhone user might use 1GB per month, an average wired user today might use as much as 20 GBs per month, and a heavy user might use as much as 200GBs a month, or about 2,000-times what the average wireless data user consumes.
This usage underscores what could be an economic problem for wireless networks: The connectivity services that users are most willing to pay for (voice, texting and email) consume low bandwidth and are available at relatively low cost. The services that demand the most network resources (streaming video and entertainment) are more costly and much less likely to be economically supported by consumers. Today, video services are supported by only 1% of mobile users. In fact, the still limited amount of video being consumed over the wireless Web is typically coming from "free" ad-supported video services, such as Hulu.com and Google's YouTube. The kicker is that these point-to-point services consume huge amounts of bandwidth, which explains why cable operators and telcos have been fighting to cap and slap extra charges against heavy users of spectrum.
Near-term, such dreary economics will prompt carriers and open networks to unload wireless broadband traffic onto wired broadband networks rather than pay to proactively increase wireless network capacity. In the future, open network service and pricing models could get interesting.
After weaving through complex (and fascinating) details, Moffett notes that the wired broadband networks are unlikely to be replaced by wireless ones and that usage levels on wireless data networks will remain relatively tight-capped. Successful networks will rely on offloading traffic from wireless networks onto existing wired networks. The big winners will be location-based search and other wireless mobile connected applications that work best on smartphones. The losers are likely to be the streaming video and entertainment applications that chew through bandwidth and are more closely associated with laptop computers.
Wireless carriers AT&T Mobility and Verizon Wireless, with high-quality 700 MHz spectrum and strong voice services, are the best-positioned to benefit from the evolution of wireless data networks, Moffett said. Wired broadband operators such as Comcast, Time Warner Cable and Cablevision also are expected to benefit. The direct broadcast satellite operators (DirecTV and Dish Network) have the most to lose if a "third pipe" alternative to the cable/telco duopoly fails to materialize. Any companies associated with wireless infrastructure will benefit under any circumstances.
Clearwire is one of the wireless data networks that currently does not have the capacity to support large-scale wireline broadband substitution. But the heralded launch of its Clear 4G Wi-Max network has the potential for a huge capacity advantage over the telcos and cable if its sponsors pay up. Clearwire reports its $2 billion underfunded--even after a $3.2 billion investment by Intel, Google, Comcast, Time Warner Cable and Bright House Networks. "It presents the tantalizing prospect of a whole new--and potentially disruptive--technology ecosystem build on low-cost Net-enabled devices, low-cost 4G wireless data networks," Moffett said.
However, there is an irony to a full-fledged alternative to the cable/telco duopoly. "The endpoint for a super-high-capacity wireless network is a wired network with a wireless node in each home, or the very configuration that already exists in many homes as a cable modem or fiber network supported by Verizon's FiOS, with a Wi-Fi router attached," Moffett adds.
That leaves us wondering: If they build it, will we come?