Down the rabbit hole

Google has quietly announced an indefinite suspension of further fiber expansion as it prepares to downsize fiber division employees and re-evaluate its fiber business model.

In a blog post tonight from Craig Barratt, senior vice president of Alphabet and CEO of Google’s Access division, it becomes clear Google is rethinking its entire fiber strategy and is likely moving towards fixed wireless technology going forward:

Now, just as any competitive business must, we have to continue not only to grow, but also stay ahead of the curve — pushing the boundaries of technology, business, and policy — to remain a leader in delivering superfast Internet. We have refined our plan going forward to achieve these objectives. It entails us making changes to focus our business and product strategy. Importantly, the plan enhances our focus on new technology and deployment methods to make superfast Internet more abundant than it is today.

Barratt outlines the immediate implications of Google’s dramatic shift:

• In the cities where we’ve launched or are under construction, our work will continue;
• For most of our “potential Fiber cities” — those where we’ve been in exploratory discussions — we’re going to pause our operations and offices while we refine our approaches. In this handful of cities that are still in an exploratory stage, and in certain related areas of our supporting operations, we’ll be reducing our employee base.

Barratt

Barratt himself is jumping ship (or was pushed). He announced in his blog entry he is “stepping away” from his CEO role, but will remain as an “adviser.”

Observing Google’s recent fiber efforts and acquisitions, it seems clear Google no longer thinks fiber-to-the-home service is an economically viable solution in light of competitors like AT&T rolling out increasing amounts of fiber and the cable industry is on the cusp of launching DOCSIS 3.1, which will dramatically boost internet speeds without a substantial capital investment.

Google’s investors have been lukewarm about the company’s economic commitments relating to its fiber broadband networks. Often built from the ground up, Google’s fiber construction complexities also include trying to navigate costly roadblocks established by their competitors (notably Comcast and AT&T), dealing with bureaucracies and red tape even in states where near-total-deregulation was supposed to make competition easy. Google Fiber has also not proved to be a runaway economic success, and now faces more challenges in light of upgrades from their competitors. Cable companies have slashed prices for customers threatening to cancel and have added free services or upgrades to persuade customers to stay, and Google’s proposition of selling consumers $70 gigabit access has proved tougher than expected.

It is highly likely the future of Google’s Access business will be deploying wireless broadband solutions powered by Webpass, a company Google acquired earlier this year. Webpass uses a high-speed point to point wireless transmission system the company claims can deliver gigabit broadband access to customers in multi-dwelling buildings and other urban areas. Webpass sells access for $60 a month (discounted to $550/yr if paid in advance) for 100Mbps-1,000Mbps speed depending on network density and capacity in the customer’s building. So far, Webpass has not been able to guarantee speed levels, and some customers report significant variability depending on their location and network demand.

Webpass’ wireless infrastructure costs a fraction of what Google has coped with building fiber to the home networks, and the installation of point-to-point wireless antennas on participating buildings has been less of a regulatory nightmare than digging up streets and yards to lay optical fiber.

But despite Webpass’ claim its performance is comparable to fiber, its inability to guarantee customers a certain speed level and its tremendous performance variability from 100 to 1,000Mbps exposes one of the weaknesses of fixed wireless networks. At a time when capacity is king, only fiber optic networks have shown a consistent ability to deliver synchronous broadband speeds that do not suffer the variability of shared networks, poor antenna placement/signal levels, or harmful interference.

There is room for wireless technology to grow and develop, as evidenced by the wireless industry’s excitement surrounding future 5G networks and their ability to offer a home broadband replacement. The emergence of 5G competition is almost certainly also a factor in Google’s decision. But even AT&T and Verizon acknowledge a robust 5G network will require a robust fiber backhaul network to support both speed and user demand. The more users sharing a network, the slower the speed for all users. No doubt Webpass has made the same assumption that cable operators did in the early days of DOCSIS 1 — current internet applications won’t tax a network enough to create a traffic logjam that would be noticed by most customers. The phone companies also learned a similar lesson trying to serve too many DSL customers from inadequate middle mile networks or traffic concentration points. (Some phone companies are still learning.)

Whether it was yesterday’s peer-to-peer file sharing or today’s online video, capacity matters. That is why fiber broadband remains the gold standard of broadband technology. Fiber is infinitely upgradable, reliable, and robust. Wireless is not, at least not yet. But technology arguments rarely matter at publicly-traded corporations that answer to Wall Street and investors, and it appears Google’s backers have had enough of Google Fiber.

Stop the Cap!’s View

At Stop the Cap!, we believe these developments further the argument broadband is an essential utility best administered for the public good and not solely as a profit-motivated venture. The path to fiber to the home service in rural, suburban, and urban communities has and will continue to come from a mix of private and public utilities, just as local public and private gas and electric companies have served this country for the last century. Where there is a business model for fiber to the home service that investors support, there is a for-profit fiber provider. Where there isn’t, now there is often no service at all. So far, the FCC in conjunction with Congress has seen fit to solve broadband availability problems by bribing private providers into offering service (usually low-speed DSL that does not even meet the FCC’s definition of broadband) with cash subsidies, tax write-offs, or occasional tax abatement schemes. Imagine if we followed that model with the nation’s public roads and highways. We would today be paying tolls or a subscription to travel down roads built and owned by a private company often financed by tax dollars.

Not every product or service needs to earn Wall Street-sized profits. Nobody needs to get rich selling water, gas, and electricity… or broadband. Public broadband networks can and should be established wherever they are needed, and they should be priced to recover their costs as well as expenses that come from support, billing, and ongoing upgrades. Naysayers like to claim municipal broadband is socialism run wild or an instant economic failure, yet the same model has provided Americans with reliable and affordable gas, electricity, and clean water for over 100 years.

Maine was made for municipal broadband.

In New York, publicly owned/municipal utilities often charge a fraction of the price charged by investor-owned utilities. In Rochester, where Stop the Cap! is headquartered, one need only ask a utility customer if they would prefer to pay the prices charged by for-profit Rochester Gas & Electric or live in a suburb where a municipal provider like Fairport Electric or Spencerport Electric offers service. RG&E has charged customers well over 10¢ a kilowatt-hour when demand peaks (along with a minimum connection charge of over $21/mo and a “bill issuance charge” of 72¢/mo). Spencerport Electric charges 2.9¢ a kilowatt-hour and a connection charge of $2.66 a month, and they issue their bills for free. There is a reason real estate listings entice potential buyers by promoting the availability of municipal utility service. The same has proven true with fiber-to-the-home broadband service.

The economic arguments predicting doom and gloom are far more wrong than right. Municipal utilities are often best positioned to offer broadband because they already have experience providing reliable service and billing and answer to the needs of their local communities. Incompetence is not an option when providing reliable clean water or electricity to millions of homes and customers have rated their public utilities far superior to private phone or cable companies.

Google’s wireless future may prove a success, but probably only in densely populated urban areas where a point-to-point wireless network can run efficiently and profitably. It offers no solution to suburban, exurban, or rural Americans still waiting for passable internet access. Clearly, Google is not the “free market” solution to America’s pervasive rural broadband problem. It’s time to redouble our efforts for public broadband solutions that don’t need a seal of approval from J.P. Morgan or Goldman Sachs.

Only half of CenturyLink’s customers in well-populated areas formerly served by Qwest can buy broadband service at 40Mbps or higher, while rural customers fare considerably worse with less than 25% able to get High Speed Internet at those speeds.

Customers have noticed and at least 65,000 canceled their broadband service with the phone company in the second quarter of this year, most presumably switching to their area’s cable operator.

“CenturyLink is by far the most abysmal telephone company I’ve ever had to deal with and I’m 63 years old,” shares Glen Canby in Arizona. Canby is a retired telephone company engineer that spent 40 years with a larger phone company serving the midwestern U.S.  “Their reviews online echo my own experiences, which have ranged from being quoted one price while being billed another, being locked into a term contract you didn’t ask for, and getting only a fraction of the speed they claim to sell.”

Canby is counted as one of CenturyLink’s 40Mbps-qualified customers, yet he actually receives less than 6Mbps service.

But that isn’t what CenturyLink tells the Federal Communications Commission. In a semi-annual broadband deployment report, the company claimed 51 percent of their customers in urban and suburban former Qwest service areas can subscribe to 40Mbps DSL or higher. But whether a customer is “qualified” to buy 40Mbps service is not the same as actually getting the speeds the company markets.

CenturyLink attempts to cover their claims with fine print attached to their FCC submission: “The numbers shown in this chart reflect the percentages of households served by DSLAMs that are capable of providing the specified broadband speeds.” (A DSLAM is a network device typically used to extend faster DSL speeds to customers by reducing the amount of copper wiring between the telephone company’s central office and the customer’s home. Customers in a neighborhood typically share space on a DSLAM, in effect sharing a single connection back to the phone company.)

“That’s clever of them, because of course the DSLAM is just one link in the chain that ends with the ‘last mile’ between that equipment and my home, and that is where CenturyLink’s phone plant is at its weakest,” Canby writes. “I spent 20 years at a phone company dealing with last-mile DSL speed issues, so they cannot fool me.”

Canby blames the condition of CenturyLink’s infrastructure between the DSLAM serving him and his home for the problems, as well as overselling DSL service by packing too many customers onto a single DSLAM.

“It might be 40Mbps service at the remote end, but it drops to around 6Mbps on a good day by the time it reaches my house,” Canby complains. “Once the sun goes down, the speed drops to 3Mbps, which is a classic case of overselling to me because too many people are trying to share one connection at the same time. It has been this way since 2008 according to my neighbors.”

Back then, phone service was provided by Qwest, the former Baby Bell providing service in 14 sparsely populated western U.S. states — Arizona, Colorado, Idaho, Iowa, Minnesota, Montana, Nebraska, New Mexico, North Dakota, Oregon, South Dakota, Utah, Washington, and Wyoming. Qwest was acquired by CenturyLink in 2011.

CenturyLink has promised to improve broadband speeds for former Qwest customers, but much of what counts as progress has been in more urban areas, while rural customers continue to languish. The company admits just 21.9 percent of rural households can get 40Mbps service. Only 47.6% can buy 12Mbps, 61.3% can get 5Mbps, and 83% can subscribe to 1.5Mbps. That leaves 17% of former Qwest customers with no broadband options at all. CenturyLink did not break out the percentage of customers that meet the FCC’s 25Mbps minimum speed definition of broadband.

“This is why CenturyLink loses customers to cable operators who have no problems trying to deliver internet access over their network, because it was built to support more bandwidth,” Canby shares. “They can usually deliver the same internet speed to customers no matter how far out they live while phone companies deal with a network built for making phone calls, not data.”

Company officials recognize they could do better and have promised investors another 2.5 million customers will be able to reach 40Mbps by the end of 2017. By the end of the year after that, CenturyLink hopes to reach 85% of customers with VDSL2, bonding, and vectoring technology to achieve 40Mbps service for most customers in their top 25 markets. But rural customers are likely to left waiting longer because of the costs to upgrade Qwest’s copper-based network, especially in smaller states like Idaho, the Dakotas and Wyoming.

“The only answer is cable or fiber broadband, and if you live in a small community it could be years before CenturyLink gets around to you,” Canby writes. “If it’s the same story all over town, I’d start advocating for a community-owned fiber network and not sit around and wait for CenturyLink to act, especially if there is no cable company in town.”

Further expansion of Google Fiber appears to be on hold as the company contemplates moving away from fiber to the home service towards a wireless platform that could provide internet access in urban areas for less money.

The Wall Street Journal today reports Google parent Alphabet, Inc., is looking to cities to share more of the costs of building faster broadband networks or using cheaper wireless technology to reach customers instead.

Six years after Google first announced it would finance the construction of fiber to the home networks, the company has made progress in wiring just six communities, many incompletely. Progress has been hampered by infrastructure complications including pole access, permitting and zoning issues, unanticipated construction costs, and according to one Wall Street analyst, the possibility of lack of enthusiasm from potential subscribers.

Google’s recent acquisition of Webpass, a company specializing in beaming internet access over fiber-connected wireless antennas between large multi-dwelling units like apartments and condos appears to be a game-changer for Google. Webpass was designed mostly to service urban and population dense areas, not suburbs or neighborhoods of single-family dwellings. Webpass’ reliance on wireless signals that travel between buildings removes the cost and complexity of installing fiber optics, something that appears to be of great interest to Google.

Google Fiber is planning a system that would use fiber for its core network but rely on wireless antennas to connect each home to the network, according to a person familiar with the plans. Alphabet chairman Eric Schmidt said at the company’s shareholder meeting in June that wireless connections can be “cheaper than digging up your garden” to lay fiber. The only question is what kind of performance can users expect on a shared wireless network. Google’s plans reportedly do not involve 5G but something closer to fixed wireless or souped-up high-speed Wi-Fi. A web video on Webpass’ website seems to concede “you get best speeds with a wired connection.”

Even Google’s wireless technology solutions provider Webpass concedes that wired broadband is faster.

Former Webpass CEO Charles Barr, now an Alphabet employee, argues wireless solves a lot of problems that fiber can bring to the table.

“Everyone who has done fiber to the home has given up because it costs way too much money and takes way too much time,” Barr said.

Barr’s statements are factually inaccurate, however. Fiber to the home projects continue in many cities, but if they are run by private companies, chances are those rollouts are limited to areas where a proven rate of return is likely. Large incumbent phone and cable companies are also contemplating some fiber rollouts, at least to those who can afford it. Many of the best prospects for fiber to the home service are customers in under-competitive markets where the phone company offers slow speed DSL and cable broadband speeds are inadequate. Rural communities served by co-ops are also prospects for fiber upgrades because those operations answer to their members, not investors. Community broadband projects run by local government or public utilities have also proven successful in many areas.

But like all publicly traded companies, Google must answer to Wall Street and their investors and some are not happy with what they see from Google Fiber. Craig Moffett from Wall Street research firm MoffettNathanson has rarely been a fan of any broadband provider other than cable operators and Google Fiber is no different.

“One can’t help but feel that all of this has the flavor of a junior science fair,” Moffett said of Google Fiber, pointing out the service has managed to attract only 53,000 cable TV customers nationwide as of December. Moffett concedes there are significantly more broadband-only customers signed up for Google, but that didn’t stop him from suggesting Google Fiber has had very little impact on increasing broadband competition across the country.

Analysts suggest Google Fiber is spending about $500 per home passed by its new fiber network. But that is a fraction of the $3,000+ per customer often spent by cable operators buying one another.

Google’s wireless deployment will likely take place in Los Angeles, Dallas, and Chicago according to people familiar with the company’s plans. Less dense cities slated for Google Fiber including San Jose and Portland, Ore., may never get any service from Google at all, but they are likely to hear something after a six month wait.

Google is also reportedly asking cities if the company can lease access on existing fiber networks. Another tactic is requesting power companies or communities build fiber networks first and then turn them over to Google to administer. The latter seems less likely, considering there are successful public broadband networks operating on their own without Google’s help.

The three candidates running for President of the Republic of China are: James Soong from the People First Party (PFP) (L), Eric Chu from the ruling Kuomintang (KMT) party (Center), and Tsai Ing-wen from the Democratic Progressive Party (DPP) (R).

Taiwan’s three presidential candidates, appearing in a nationwide debate on Sunday, promised to deliver improved High Speed Internet in the Republic of China, with some candidates committing to give broadband away for free to low and middle-income families.

Taiwan is making broadband expansion and improvement a top national priority, as the country races towards delivering gigabit wired broadband and 5G wireless service. The government wants to boost the country’s broadband ranking, now 33rd in the world.

Bringing speeds up while reducing broadband bills is the goal of Eric Chu, the candidate from the ruling Kuomintang (KMT) party. During his terms as leader of New Taipei and Taoyuan County, Chu presided over a major expansion of Internet penetration rates. Chu believes the next step is to make broadband service free of charge for low/middle-income residents and deliver nationwide free Wi-Fi to every centimeter of Taiwan.

Internet providers would still profit from selling faster access to customers willing to pay for it, but Chu’s policies continue a theme that broadband access is a basic human right, a position increasingly popular in the country. Voters appeared skeptical of Chu’s claims, however, because the KMT has garnered a reputation of being in bed with big business during its last two terms in office. But that has not stopped Chu from criticizing telecom executives for not doing more to invest and eventually offer next generation 5G wireless service in Taiwan.

James Soong, from the People First Party — considered to have a close (but frequently tense) alliance with the KMT  — predictably agreed with Chu, but also wants Taiwan to do more to protect Internet privacy and online safety. Soong wants to completely scrap the country’s legacy copper wire telecommunications infrastructure and replace it with fiber optics, delivering fiber service to every home and business in Taiwan. With a fiber upgrade, Soong is convinced Taiwan will achieve his goal of top-10 broadband status.

Tsai Ing-wen from the Democratic Progressive Party (DPP) said when private companies don’t deliver, it is government’s responsibility to address the digital divide, by making high quality service affordable and fast. Taiwan’s telecom companies are paying close attention to the DPP candidate because polls make her the favorite to become the next president of Taiwan, after the election on Jan. 16.

“The use of broadband Internet service should be part of the people’s basic human rights,” she said. “It is also important to narrow the digital divide to improve educational opportunities for children in remote areas and develop children’s digital capabilities.”

With broadband being treated as a high priority issue in the presidential race, Taiwan’s largest broadband provider, Chunghwa Telecom – 中華電信, has announced the first commercial deployments of G.fast technology – the newest generation of DSL – across Taiwan.

Israeli chipmaker Sckipio demonstrated G.Fast technology at CES 2016 in Las Vegas this week, claiming it is faster than traditional DSL and cable broadband. In a limited demonstration, the company demonstrated download speeds achieving 750Mbps over traditional copper wire networks, about 50 times faster than average broadband speeds. Sckipio promised G.Fast technology will debut in the United States later this year.

Frontier Communications has topped AT&T’s penchant for grandiose Fiber to the Press Release announcements with a new gigabit fiber to the home service now being promoted in Connecticut, despite being available to only two homes in a single upscale subdivision in North Haven.

Frontier FiberHouse is Frontier’s answer to Verizon FiOS, says Joseph Ferraiolo, Frontier’s regional general manager in New Haven County. Ferraiolo told the New Haven Register Frontier has introduced the service to a pair of homes in Lexington Gardens, a new single-family subdivision.

Frontier’s expansion of the service in 2016 does not appear to be exactly aggressive, with plans to only wire up to 200 newly built homes in the immediate area.

Frontier’s fiber network relies on a Gigabit Passive Optical Network (GPON) and is intended to replace copper telephone wiring.

Ferraiolo admits Frontier is currently favoring new housing developments where fiber can be dropped in a conduit/pre-existing trench during construction without the cost of tearing up yards and streets. But he also claims Frontier will make a commitment to any municipality that gets the fiber service that it will be available to every part of the community, not just those likely to be most profitable. If Frontier keeps its promise, it will be the first time the phone company has provided customers with universal access to uniformly high-speed broadband. Even its acquired FiOS networks in Indiana and the Pacific Northwest are not guaranteed to be available to every resident.

“We think this is a good option for us: new builds, small complexes,” Ferraiolo said. “The developer is very happy with it and we’re very happy with it.”

Customers like William Morico will believe it when they see it.

“We have been trying to get ‘high-speed’ Internet in our neighborhood for years, well before the Frontier disaster,” Morico writes. “All we want is the 12-18Mbps service that is advertised and available elsewhere in New Haven. [We] cannot get any answers from Frontier. Even their customer service and tech staff are frustrated with this company. It’s time for the state gig project.”

The company claims it is “exploring” other rollouts of Frontier FiberHouse in Stamford and New Haven, but there are no specifics.

Some observers question the timing of Frontier’s fiber announcement, noting state and local officials are still considering a private-public partnership that could lead to a public statewide gigabit fiber network in Connecticut. News that a private company is willing to shoulder the entire expense of a fiber project could be used in legislative efforts to derail Connecticut’s CT Gig Project. But Frontier has offered no guarantees whether or if it intends to blanket its service area across the state with fiber or limit FiberHouse to a de-facto demonstration project in a handful of homes in new housing developments.