Google to Build Fiber to Home Test Networks

In the latest version of its Internet access demonstration projects, Google says it will build some wholesale fiber-to-the-home networks. James Kelly, Google project manager, says Google is looking for communities interested in becoming trial sites, and will be accepting requests for information until March 26, 2010.

Initial plans call for building FTTH facilities serving 50,000 people, perhaps as many as 500,000. That's a bit indefinite. Assume a typical household has about 2.5 people in it and one could see perhaps 20,000 to 200,000 homes connected.

But there are lots of unanswered questions. It isn't clear whether Google means "a network passing X number of homes" or "a network serving X number of homes." Those are very different sorts of numbers if Google builds anyplace where a strong cable operator and a strong telco provider already are in business.

As a demonstration project, Google would learn as much serving 20,000 potential customers as 200,000. Frankly, there is very little that is unknown about the cost of building a fiber-to-home network, really. What Google might be interested in is the business case for a wholesale broadband network that didn't have to supply voice and traditional video services.

But again, there is very little that is unknown about that business case, either. We know the cost to build the network and operate it as an ISP. You can derive commercial rates by looking at what is charged in the local, regional or national markets, multiply by take rate of homes passed and you'd have your scenario without digging a single trench.

Kelly says "we are going to try out new ways to build and operate fiber networks and share what we learn with the world."

There likely is not much to be uncovered about the process of building FTTH, so Google likely means sharing what it learns about the business model for wholesale networks, which a few have experimented with in the U.S. market.

Broadweave, which operates in Provo, Utah, is an example, though it is unclear whether the firm will continue to allow wholesale customers to use its network. The current thinking seems to be to operate as a typical retail triple-play provider, rather than as a wholesale provider of access to third parties who may wish to do so.

Broadweave Networks took over a project originally started as a municipal FTTH network, paying $40.6 million for the assets, and struggled in 2009. The company drastically slowed its growth among residential customers to save on the high cost of new sign-ups in the spring of 2009, reports the Daily Herald.

Broadweave began asking prospective customers to foot part of the $1,000 installation bill in an effort to discourage new customers and effectively slow growth while the company grapples with ongoing financial concerns.

The company is instead decided to focus on commercial accounts and reactivation of former customers who already have optical drops in place.

The flow of new customers reportedly was slowed from hundreds a month to "tens or dozens" a month.

Google might attempt to prove the thesis that an FTTH wholesaler can make money strictly as a supplier of access services to third party partners who simply lease capacity on such networks. The big question always has been whether any such network actually is viable in markets where there are strong cable and telco competitors already in place.

Google previously has built limited municipal Wi-Fi networks as well. It isn't clear what Google believes it might have learned from those experiments.

In truth, the gambit most likely is simply another tool to be used in Google's lobbying for setting of national broadband policies, and not much more.