Google Suspends Several Google Fiber Builds; Fixed Wireless the Reason?

It is too early for those of us outside Google to figure out what it means that Google Fiber has halted or suspended its plans to build Google Fiber networks in San Jose, Calif., Palo Alto, Calif. and Mountain View, Calif.

The stated reason for the pause is to allow Google Fiber to explore whether fixed wireless is a better option for building those networks. Construction speed and cost as well as lower overall capital investment are potentially major advantages.

Google apparently believes its gigabit speed offer can be offered using fixed wireless, instead of optical fiber.

Some skeptics might argue that the whole purpose of Google Fiber was to goad other competitors into upgrading to gigabit speeds--something that is happening nationally--and never to become a major ISP in its own right.

Whatever the outcome, it now is becoming clearer that physical media choices are increasing, where it comes to gigabit networks. In addition to fiber to the home, cable operators are doing so with their hybrid fiber coax networks, and it now appears fixed wireless and mobile networks (5G) will be able to do so, as well.

To an astounding degree, that is a major change in platform capabilities and potential business models. Until recently, it was generally assumed that only fiber to the home could support commercial and widespread gigabit speeds.

That will be less the case in the coming years, as 5G networks are deployed--in fixed and mobile variants--and as much as 29 GHz of new communications spectrum, including at least 7 GHz of unlicensed spectrum is released for use in the U.S. market.

AT&T says it has achieved speeds up to 14 Gbps using millimeter wave radio in what appears to be a point-to-point application, and speeds up to 5 Gbps to two users, in what appears to be point-to-multipoint application.

That test appears to have used 15-GHz frequencies. AT&T says it now will test propagation at 28 GHz.

Separately, Google has asked the Federal Communications Commission for authorization to conduct radio experiments in the new Citizens Broadband Radio Service  (CBRS) band, at 24 U.S. locations.

That is important for several reasons. First, the CBRS is the first U.S. frequency band to feature shared spectrum access: commercial users and licensed government users will share access to bandwidth.

Second, CBRS will be a major new way for Google--and other ISPs--to provide Internet access services, beyond Google Fiber.

Third, the move suggests the coming important role of fixed wireless in the U.S. ISP business.

Google plans to deploy initially in Atwater, Calif., Mountain View, Calif., Palo Alto, Calif., San Bruno, Calif., San Francisco, San Jose, Calif., Boulder, Colo., Kansas City, Kan., Omaha, Neb., Raleigh, N.C., Provo, Utah, and Reston, Va.

Those locations skew heavily to major urban areas near Google’s headquarters, some sites where Google Fiber already operates, but also some new smaller-market locations.

The initial test locations also indicate Google wants to test interference issues in areas where licensed users are active (coastal regions are issue for some licensees).

Google apparently also is looking at locations where it already operates Google Fiber, potentially adding a new access technology option to the current fiber-to-home approach.

Google says “operations vary from 7 km to 40 km from the geographic center point of each test area.” That implies potential testing of signal propagation and interference testing ranging from four miles to nearly 25 miles.

The test locations are not commercial launch sites, Google says.

AT&T Demos 14 Gbps at 15 GHz

AT&T says it has achieved speeds up to 14 Gbps using millimeter wave radio in what appears to be a point-to-point application, and speeds up to 5 Gbps to two users, in what appears to be point-to-multipoint application.

That test appears to have used 15-GHz frequencies. AT&T says it now will test propagation at 28 GHz.

Separately, Google has asked the Federal Communications Commission for authorization to conduct radio experiments in the new Citizens Broadband Radio Service  (CBRS) band, at 24 U.S. locations.

That is important for several reasons. First, the CBRS is the first U.S. frequency band to feature shared spectrum access: commercial users and licensed government users will share access to bandwidth.

Second, CBRS will be a major new way for Google--and other ISPs--to provide Internet access services, beyond Google Fiber.

Third, the move suggests the coming important role of fixed wireless in the U.S. ISP business.

Google plans to deploy initially in Atwater, Calif., Mountain View, Calif., Palo Alto, Calif., San Bruno, Calif., San Francisco, San Jose, Calif., Boulder, Colo., Kansas City, Kan., Omaha, Neb., Raleigh, N.C., Provo, Utah, and Reston, Va.

Those locations skew heavily to major urban areas near Google’s headquarters, some sites where Google Fiber already operates, but also some new smaller-market locations.

The initial test locations also indicate Google wants to test interference issues in areas where licensed users are active (coastal regions are issue for some licensees).

Google apparently also is looking at locations where it already operates Google Fiber, potentially adding a new access technology option to the current fiber-to-home approach.

Google says “operations vary from 7 km to 40 km from the geographic center point of each test area.” That implies potential testing of signal propagation and interference testing ranging from four miles to nearly 25 miles.

The test locations are not commercial launch sites, Google says.

For Ting, Operating Costs are Key to Business Model

Low overhead, and low operating costs--rather than any special capital investment advantages-- appear to be a key element of the business plan for Tucows mobile and Internet access businesses.

But picking the right market might be even more important. A delay in Google Fiber construction plans in Portland, Ore. might provide an example. Where Google Fiber might once have hoped to be the only provider of gigabit Internet access in Portland, both major suppliers Comcast and CenturyLink already are moving to do so.

No matter how attractive an offer Google Fiber might supply, it could be only the latest of three major ISPs to offer gigabit services in that market.

And that likely will be the case for any successful new facilities-based independent Internet service providers, as well. That has been the case for cable TV operators, who now are the market leaders for high speed Internet access in the U.S. market, and likely is true for Google Fiber.

New gigabit Internet access operations, built using fiber to home network platforms, probably cannot significantly outperform tier-one providers on network build costs. So to make the business case work, assuming use of price discounts as a key marketing tool, other costs must be sharply lower.

In its mobile business, Ting seems able to acquire new customers for less than a $100, far lower than one sees at larger service providers, where costs can range from $300 and up.

For Tucows, that is key, as it does not seem able to build fiber to home networks cheaper than might be expected for a tier-one service provider.

Though it estimates an average cost of $2,500 per customer to build FTTH plant, Ting Internet is spending closer to $3,000 per customer for its Holly Springs, N.C. network.

At an assumed 20 percent take rate in the first year and 50 percent in five years, that implies a per-passing cost of about $600. That is within the range of current FTTH network build costs, though arguably not including activation costs.

And, so far, Tucows seems to generating gross margin for network access of 48 percent. And, in large part, that seems to be the result of lower operating costs. Some telcos have operating costs in the 60-percent range.

As a percentage of revenue, total operating expenses were down about 0.5 percentage point to 20.4 percent,  compared to a year ago, Tucows says.

At the same time, the major carrier price reductions for mobile data have narrowed Ting’s competitive positioning. “Over the past couple of years as the carriers have reduced data prices, we have clearly lost some of our competitive positioning,” said Elliot Noss, Tucows CEO.

That further emphasizes the key role played by operating costs in the business model. Ting operates with lower retail prices and typical capital investment. Only in the marketing and other operating parts of its business can Ting gain advantages.

Amazon Web Services Chooses to Build Rather than Buy

Amazon has become a part owner of a new trans-Pacific submarine cable network intended to connect data centers in the United States, New Zealand and Australia.

When the Hawaiki Submarine Cable comes online in 2018, it will provide considerably more bandwidth between the US, Australia, and New Zealand than available today, and is expected to lower Amazon Web Services transmission costs.

Otherwise, on those routes, AWS would be buying capacity from Telstra Endeavour and Southern Cross Cable Network.

Enterprises often have “build versus buy” choices where it comes to communications infrastructure.

Very-small firms might always find that buying communications services makes more sense than building and owning infrastructure. Think of phone systems. Still, at some point, owning a phone system is cost neutral.

For large enterprises, owning phone systems traditionally has been cheaper than leasing services from telecom providers.

Since the 1980s, U.S. enterprises seemingly have vacillated between building their own private networks and leasing capacity from telecom service providers. Of course, cloud services providers have unusually important connectivity requirements, making a “build” strategy more valuable. That is why Google, AWS and Microsoft operate their own data centers, for example.

In the era of cloud services, at least some enterprises are finding that owning and operating their own infrastructure is, in substantial part, a reasonable approach.

Price Controls on Special Access Will Depress Optical Facilities Investment in Rural Areas, Economist Argues

Price controls on special access services  likely would depress investment in new optical facilities for business customers in rural areas of the United States, says James E. Prieger Professor of Economics and Public Policy Pepperdine University School of Public Policy.

In some cases, price controls would reduce retail prices as much as 22 percent to 32 percent, FCC economists have predicted.

Dr. Hal Singer argues that a 30 percent decline in revenue from extending special access regulations to ILECs’ fiber networks would lead to an estimated 55 percent decline in investment in business broadband, for example.

Special access is a product no different than rental apartments, in one clear sense. Rent control policies do keep retail prices low, but also are a deterrent to construction of new affordable housing.

Most-if not all--economists would agree, based on principle and historical behavior of firms in competitive markets with price controls, that price controls will reduce investment.

That might be good for renters occupying price controlled apartments. But new affordable rental housing is discouraged. In fact, even routine maintenance on a rent-controlled unit will fall.

Special access service is based on legacy circuit switched network technology destined to be replaced by Ethernet-based services. So no large service provider has much incentive to invest heavily in additional special access facilities, compared to Ethernet networks, under the best of circumstances.

At the same time, lower potential profit also will deter potential competitors from investing in their own facilities. In substantial part, the same business case issues that make additional special access investment unattractive for tier-one providers also will apply to would-be facilities-based providers as well.

As always is the case, abnormally high profits create incentives for new entrants to enter a market. Low profits are a discouragement.

Price controls, as always, will depress potential return from new special access facilities even further. The business case, as always, is tougher in rural areas than in urban areas.

WOW Launching Gigabit Internet Access Services

WOW! Internet, Cable & Phone, a provider of triple play services, is launching broadband speeds up to one gigabit per second in five U.S. markets by the end of 2016.

The rollout will provide gigabit speeds to thousands of residential and business customers in Huntsville and Auburn, AL; Evansville, IN; Knoxville, TN; and Grosse Pointe Shores, MI.

In Evansville and Auburn, the firm  will be the first service provider to bring residents and small businesses the new gigabit service. It might be noteworthy that the gigabit launch is supplied both by hybrid fiber coax and fiber to home facilities.

These days, in the fixed network Internet access market, access platform matters less than headline speed.

The company provides Internet access and triple play service in 20 markets, primarily in the Midwest and Southeast, including Illinois, Michigan, Indiana, Ohio, Kansas, Maryland, Alabama, Tennessee, South Carolina, Florida and Georgia.

At December 31, 2015, the firm’s networks passed three million homes and served 778,000 total customers, reflecting a total customer penetration rate of 26 percent.

The company typically operates in affluent suburban communities and have a customer base with income levels above the national average, the company says.

Few U.S. Business ISP Customers Would Recommend That Others Buy Their Chosen Services

source: PC Magazine

“Value” is what buyers say it is, despite all efforts by suppliers to define and shape perceptions of product value. So customer surveys often are important measures of value perceptions. One recent survey of business customers of Internet service providers found RCN and Verizon FiOS at the top of the rankings, Charter Communications ranked above average but all other surveyed major providers below average among business customers.

For those who rely on “net promoter scores” (a measure of customer loyalty based on the answer to a single question: “How likely is it that you would recommend our company/product/service to a friend or colleague?”), the results are consistent with overall rankings.

RCN and Verizon are at the top; Charter and Cox Communications are in positive territory, but all others have negative net promoter scores.

The “likelihood to recommend” scores for U-Verse, TWC, Comcast, and CenturyLink are all so low that the resulting Net Promoter Scores are all negative numbers.

Broadly, those rankings are consistent with polls of consumer satisfaction with U.S. Internet service providers. According to the American Consumer Satisfaction Index, Verizon is significantly ahead of other providers.

So far, no major entity seems to track Google Fiber performance as a routine part of such indices, so it impossible to say how consumers rank Google Fiber.

But the ISP industry is the lowest-ranked industry studied as part of the ACSI program. Linear TV fares only slightly better than the ISP industry.