Behind Google's Broadband Strategy

Search Giant's Fiber-To-The-Home Plan More Political Posturing Than Bona Fide Business

Is Google planning pricey Potemkin broadband villages?

As part of an expensive lobbying move, the search giant last week unveiled plans to fund fiber-to-the-home broadband networks in the U.S. that would deliver unprecedented speeds of 1 Gigabit per second. The search giant said it initially plans to deliver such lightning-fast speeds to at least 50,000 customers, at what Google promised would be a “competitive” price — although it didn't say specifically what that meant — with the goal of eventually providing service to 500,000 subscribers.

The news that one of the wealthiest technology companies is jumping into the broadband business would, at first glance, seemed to strike fear in the hearts of cable and telco distributors afraid of another “triple-play” provider. After all, the search-engine giant offers Google Voice phone service and video with YouTube (although the latter has, to date, been an unprofitable business for Google). Wall Street, however, dismissed Google's “Think Big With a Gig” campaign as a stunt.

“The Google plan is short on details, with no information on capital spending, and, in our view, should primarily be seen through the lens of regulatory posturing,” Sanford Bernstein senior analyst Craig Moffett wrote in a research report. “We do not view Google's announcement as a serious threat to the broadband businesses of either the cable or telecom operators.”

Indeed, the price tag for building out such “experimental” über-networks in undetermined locations could easily run into tens of millions of dollars. Google didn't say when it would initiate the project or how much it intended to spend, but analysts estimated the cost of a FTTH buildout to be at least $1,500 per subscriber and potentially far greater.

Google's stated aim is to push the Federal Communications Commission to require higher speeds and open networks as part of the agency's broadband plan, due to Congress on March 17.

“We've urged the FCC to look at new and creative ways to get there in its National Broadband Plan, and today we're announcing an experiment of our own,” read a post on Google's corporate blog that was attributed to company product managers Minnie Ingersoll and James Kelly.

It's unclear whether the gambit will have the desired effect of influencing the government's broadband initiatives and establishing network-neutrality laws.

According to Google, the company is essentially looking to demonstrate the type of network that it has argued is needed to handle the future of bandwidth-hungry video applications. The company offered such blue-sky examples as “collaborating with classmates around the world while watching live 3D video of a university lecture.”

Cable's response? The National Cable & Telecommunications Association emphasized that the cable industry has invested $161 billion over the past 13 years to build a broadband infrastructure that is available to 92% of U.S. homes.

“We look forward to learning more about Google's broadband experiment in the handful of trial locations they are planning,” NCTA vice president of communications Brian Dietz said.

Cable operators, Dietz continued, will “continue to invest billions more to continually improve the speed and performance of our networks and provide tens of millions of consumers with the best possible broadband experience.”

Privately, many cable operators say they see the move as a ploy to gain regulatory favor. Others pointed to Google's track record for making broad statements and not fulfilling them, such as its 2006 plan to build a municipal Wi-Fi network in San Francisco and other cities.

“They built out the city of Mountain View, [Calif.], which consists of the Google complex and not much else; 20,000 people use their tremendous Wi-Fi experiment,” said a cable executive from one of the five largest operators.

Still, some MSOs are not taking the notion of Google broadband lightly. One midsized operator who asked not to be named said that although the cost of building a fiber network appears onerous, Google could get around those costs by building in high-density areas with the bulk of infrastructure already in place.

“I would be surprised if the communities they pick are ones where the average cost to build is north of the [$1,500] average,” said the cable executive, who asked not to be named. “Density will be a big factor.”

Meanwhile, Verizon Communications, which has spent north of $18 billion on its FiOS fiber-to-the-home network, didn't betray any fear of the Googleplex, at least publicly. “The Internet ecosystem is dynamic and competitive, and it's delivering great benefits to consumers,” Verizon executive vice president of public affairs, policy and communications Tom Tauke said. “Google's expansion of its networks to enter the access market is another new paragraph in this exciting story.”

Moffett noted that Google doesn't plan to offer cable TV service over the network, nor will it likely be profitable. “In essence, Google appears to be promising Verizon FiOS with much higher costs and a fraction of the revenues.”

FCC chairman Julius Genachowski, though, responded favorably to Google's announcement. “This significant trial will provide an American test bed for the next generation of innovative, high-speed Internet apps, devices and services,” he said in a statement. “The FCC's National Broadband Plan will build upon such private-sector initiatives and will include recommendations for facilitating and accelerating greater investment in broadband, creating jobs and increasing America's global competitiveness.”

Also praising the Internet company's proposal was Sen. John Kerry (D-Mass.), chairman of the Senate Commerce Committee, who said, “Google's announcement today amounts to a nationwide competition for communities to step up and make the case for what a next-generation network could do for them and then show America what is possible.”

Still unclear at this point is how a single household will be able to take advantage of anywhere close to the 1-Gbps speed Google is promising. Google boasted that the FTTH service would be 100 times faster than a typical 10-Mbps cable broadband connection, appearing to imply that cable and telco providers are unable to deliver such speeds.

True enough, even with the channel-bonding capabilities of DOCSIS 3.0, cable providers would need to free up substantial spectrum on its hybrid fiber-coax networks to provide 1 Gbps to subscribers. Comcast, Cox Communications, Time Warner Cable and others are delivering 50-Mbps tiers today, and Cablevision Systems and Mediacom Communications are topping 100 Mbps.

Verizon's residential FiOS Internet carries a top commercial speed of 50 Mbps, but the telco's FTTH network is able to deliver 2.5 Gbps to each home, combining TV and Internet, and, as configured, could deliver Internet download speeds of 400 Mbps. The telco has tested 100-Mbps connections over FiOS and also has successfully tested a passive optical network system, known as XG-PON, which can transmit data at 10 Gbps downstream and 2.4 Gbps upstream.

Today, though, there aren't any practical Internet applications that would be able to use even 5% of a pipe the size that Google has proposed to offer. Cable and telco operators say there aren't many consumers that find a need for a 50-Mbps pipe, much less a 100-Mbps or 1-Gbps one, although businesses would be able to use such fast connections to share among multiple employees.

The plan could be costly for Google to fully realize. Building an FTTH network to deliver 1-Gbps Internet access to 500,000 subscribers would cost $750 million or more, according to Will Richmond, a media-industry consultant and publisher of VideoNuze.

“Even for Google, that's a very big number, especially considering the company has said it has no intention of actually pursuing this as a business,” Richmond said.

Richmond arrived at his estimate based on total costs of Verizon's FiOS project, of roughly $1,000 per home passed and another $500 per subscriber connected.

Google's costs could be much higher, Richmond pointed out, since homes passed would by necessity be higher than the percentage of an area's residents who would opt for the experimental service. If Google's FTTH penetration rate was 15%, a decent rate for a new entrant to achieve within a few years, Google may need to pass fiber by 10 homes for every one it gets as a participant in its experiment, he said.

For now, Google has simply issued a request for information from municipalities across the country that would like to have the network built there, with a March 26 deadline for responses. The company said it will provide lessons learned from building fiber networks and, in a nod to its support for network neutrality, an “open-access” network with a choice of multiple providers managed in an “open, nondiscriminatory and transparent way.”

Google was not talking specifics on the network trials. “Our priority is to first identify interested community partners,” Google spokesman Dan Martin said.

As to the suggestion that the broadband service is a lobbying gambit, Martin said the purpose was certainly to push others to follow suit.

“Like our municipal Wi-Fi network in Mountain View, the purpose of our 1-Gigabit fiber-to-the-home project is to experiment and learn,” he said. “Google benefits when the Internet gets better, faster and more accessible, and we hope that our innovation and experimentation in these limited, trial deployments will inform and accelerate ultra-high-speed deployments elsewhere.”

Mike Farrell contributed to this report.

Google has made feints before on Internet access:

FCC's 700-MHz auction: Google in 2007 lobbied the FCC to adopt a rule requiring the winning bidder of some of the most desirable DTV spectrum to provide “open access” to third-party applications and devices if a minimum bid price of $4.6 billion were met. Google bid to ensure that price was reached, then dropped out of the auction.

Wi-Fi access in San Francisco: Google aimed to provide free wireless service in the city in partnership with EarthLink, but the project was abandoned because of concerns over EarthLink's financial viability. Google provides free Wi-Fi only in Mountain View, Calif., where it is based.