West Virginia broadband customers see slower-than-advertised speeds
CHARLESTON, W.Va. — The slow speed of broadband Internet and lack of availability in rural areas in West Virginia has been well-documented, but according to new data released by a broadband metrics company, many West Virginians who do have broadband aren’t getting advertised speeds.
Ookla, which runs the popular broadband speed test website Speedtest.net, says on average, residential Internet subscribers in Huntington are only receiving 59.4 percent of the download speeds they’re paying for, which ranks 795th out of 800 U.S. cities tested in the past year.
West Virginians receive 81.6 percent of advertised Internet speeds, on average — the second-worst figure in the nation.
The data, first published by The Wall Street Journal, was compiled from tens of millions of Speedtest.net tests as well as surveys of 646,404 Speedtest users’ subscribed Internet speeds over the past 12 months.
Jamie Steven, chief marketing officer for Ookla, said data on users’ advertised Internet speeds is collected from voluntary surveys users may take after performing a speed test. The survey asks users what their advertised upload and download speeds are, how much they pay for Internet service and who they purchase Internet service from, among other questions.
Steven said the data included speed tests from cable, DSL and satellite Internet subscribers, and excluded speed tests performed over mobile data connections.
“We have so many of those surveys, we ensure the quality of the data,” Steven said.
The problem isn’t isolated to Huntington. The four other West Virginia cities listed in Ookla’s 800-city report fall in the bottom 10 percent of the list.
Martinsburg residents receive 74.93 percent of advertised download speeds (770/800), followed by Parkersburg at 76.75 percent (764/800), Morgantown at 79.05 percent (756/800) and Charleston at 80.43 percent (750/800).
Huntington Mayor Steve Williams publicly voiced his disappointment with Comcast during the April city council meeting. Williams said he has met with Paul Comes, director of governmental affairs for Comcast, and told him Huntington residents deserved “world-class” services. Williams said Comes gave “every indication” that the company would provide the level of service Huntington is capable of receiving.
“They are aware of what we’re trying to do and as we become more specific as to what our needs are, we’ll see whether Frontier or Comcast are able to provide those needs,” Williams said. “If not, we’ll seek partnerships to develop that on our own.”
Nationally, Frontier Communications customers receive 80 percent of the download speeds they pay for on average, while Comcast customers receive 98 percent of advertised download speeds, according to Ookla. These numbers fall roughly in line with data in a report published by the Federal Communications Commission in February 2013, which specifically noted that Frontier improved its performance 13 percent between the July 2012 and February 2013 reports.
Comcast and Frontier are the two prominent residential high-speed Internet service providers in Huntington and serve customers in many parts of West Virginia.
Williams said the city of Huntington has two options it can explore if service doesn’t improve. It could negotiate with another provider to buy out Comcast’s existing infrastructure — Williams said he has already talked with some other providers, but no action has been taken.
Alternatively, Williams said the city could look to build its own data network, separate from any service provider. The city of Santa Monica, Calif. has won numerous awards for its municipal fiber optic data network that has spurred economic activity and driven down broadband costs in the city. The city network gives businesses affordable access to Internet speeds of up to 10 gigabits per second and is the backbone for a free citywide Wi-Fi network.
Huntington’s franchise agreement with Comcast expired in 2008, but City of Huntington communications director Bryan Chambers told The Herald-Dispatch common law dictates that if both sides continue as if there is still an agreement, then courts view the agreement as still valid.
That has happened so far, but Williams said that may change if Comcast cannot improve its television and Internet services in Huntington.
Williams said he knows of at least one business that has left Huntington due to a lack of sufficient Internet service. He said the business, which did technical work in the motion picture industry, moved to North Carolina last year because “they couldn’t get the speed of broadband they needed at an affordable price.”
“We operate under the philosophy ‘world-class, or nothing at all,’” Williams said. “What we have right now, we have settled. Whether it be Huntington or any city in West Virginia, we should expect nothing less than world-class service.
“Nothing was resolved in the meeting (with Comcast) other than (Comes) understood where our concerns were,” Williams said. “We know (Comcast is) a world–class organization, and just because we might be the size of a gnat on an elephant’s butt, it’s still pretty significant when it relates to 50,000 residents in and around Huntington.”
The argument against speed test data
Representatives from Frontier Communications and Comcast were contacted and asked about the data, but neither company had clear answers as to why customers in Huntington and across West Virginia fall so far behind in obtaining advertised Internet speeds.
Frontier Communications spokesman Dan Page said while Ookla speed test data is collected accurately, there are many factors that can affect the speeds reported. He said the age of networking equipment in a customer’s home, the wireless signal level of a computer running a speed test over a Wi-Fi connection, and the number of other devices on a customer’s network that may be consuming bandwidth during a speed test are all factors that could affect speed test results.
“For example, if you have two kids on Xbox and someone streaming video, that has a lot of demand on bandwidth, and that will affect someone using a PC for email,” Page said. “Netflix and those kind of services really have require a lot of bandwidth, and we’ve seen that growing over the years. It would have bearing on the speed you’re able to see.”
Page also said users who take a survey after performing a speed test may not actually know what level of Internet service they pay for and may enter incorrect numbers, which could skew data.
Steven acknowledged that other devices consuming bandwidth on a home network can skew results, but said even most older networking equipment is capable of handling today’s Internet speeds.
Page noted that Frontier is aggressively building out its DSL Internet infrastructure in West Virginia. He said the company has upgraded 39 local broadband distribution sites in the Huntington area since July 2010 and that an additional 23 site upgrades are planned.
Page said Huntington residential customers can subscribe to up to 24 megabits per second Internet access, but that language Frontier uses in advertising and contracts indicates customers will get speeds “up to” a certain level, and that customers are not necessarily guaranteed to receive a particular speed at all times. Jim Martin, President and CEO of CityNet and a member of the West Virginia Broadband Deployment Council board, said Internet customers have no legal recourse if their ISP doesn’t provide advertised speeds; their best option is to change providers.
“We do provide language to our customers that we provide up to certain speeds, and we are enhancing speeds as we can,” Page said.
Though Page said user-performed speed tests are prone to being affected by several different factors, Frontier contracts with Ookla to host speed test servers throughout the country, including two in West Virginia, to collect performance data from its customers. Comcast also hosts speed test servers that are licensed through Ookla.
The middle mile
Steven said there are many factors that can affect the speed of Internet service in a given area. He said rural areas typically see lower speeds than advertised because service often degrades further away from a cable or phone company’s central office (CO) or head-end over last-mile networks.
“If you’re talking about more distance and less-dense areas, it can be more difficult to provide high-performance service,” Steven said. “You may have more customers on the fringes of your network and that might affect your performance. You might have customers within a mile, but with a high proportion of customers 10 miles away or 5 miles away, it gets harder and harder for that service to reach its potential.”
Steven said competition between ISPs in a given area can also drive better Internet performance. Areas with more ISPs competing against each other are more likely to have faster Internet speeds and better service quality.
For example, broadband prices from competing ISPs in Santa Monica, Calif. reportedly went down 20 percent after the city linked its fiber network to a middle-mile, open-access network in Los Angeles and began offering competitive prices for high-speed connections that serve as a vital link between the city’s motion picture industry and the major studios in Hollywood. The city was able to lease a fiber optic cable from a competing ISP between the city and a major data center in Los Angeles, where bandwidth could be purchased more cheaply.
Martin and Steven said lower-than-advertised broadband speeds in West Virginia may be connected to a lack of competition between ISPs in West Virginia, especially in rural areas.
“If you look at any market where you have competing broadband competitors, you usually find customers are getting a better price for a better product, better service,” Martin said.
According to data from the West Virginia Broadband Mapping Program provided to the Daily Mail, 43,314 of about 1.1 million “addressable structures” in West Virginia have access to broadband Internet (at least 3 mbps download speeds and 1 mbps upload speeds) through only one provider — either a wireline or mobile broadband provider. 13,408 of those structures only have access to mobile broadband through one cellular data provider, and do not have access from a traditional, wireline provider.
Tony Simental, state geographic information systems coordinator and West Virginia Broadband Mapping Program coordinator, said “addressable structures” includes residences, businesses and uninhabited buildings, such as railroad outbuildings and buildings at the bases of towers.
The lack of competition in rural areas, Martin said, can be attributed to the state’s lack of middle-mile, open-access networks — super-high-bandwidth Internet pipelines that ISPs can purchase bandwidth from at wholesale rates, then re-sell to customers. Martin said these middle-mile backbone connections run through West Virginia’s major cities — like Wheeling, Charleston and Huntington — but haven’t made it to the state’s more rural, underserved areas.
West Virginia received $126.3 million in federal Broadband Telecommunications Opportunities Program (BTOP) funds in 2010 for its “Statewide Broadband Infrastructure Project” — a plan that called for adding 2,400 miles of this open-access, middle-mile fiber that would link together 1,500 “anchor institutions,” including schools, libraries, government offices and public safety agencies, on a statewide, high-speed network. Local ISPs would also be able to tap into the open network for bandwidth.
Frontier Communications was awarded $42 million by the state to run 675 of fiber to anchor institutions, but an independent consulting firm hired by the state in 2013 found the state had created an “unintended monopoly” and “unusable network except for Frontier,” according to the Charleston Gazette.
The consulting firm found that Frontier had not built a middle-mile, open-access network, but instead built last-mile fiber strands from anchor institutions to telephone poles, where the fiber connected to Frontier’s private network.
“As designed, the fiber extensions only connect the community anchor institutions to a private network that does not provide equal opportunity access,” the consultant’s report says. “Without interconnection agreements, the network is not functional for other providers, the state or other users.”
Martin said the city of Philippi in Barbour County serves as a prime example of innovation suppressed by the lack of a middle-mile, open-access network. The city contracted Cable Constructors, Inc. and Motorola in 2005 to build a fiber-to-the-home network to the city’s residents and outlying areas. Fiber-to-the-home networks can carry a virtually unlimited amount of Internet bandwidth, along with voice, video and other services.
The problem, Martin said, is that without access to a middle-mile, open-access network for the city’s fiber network to interconnect with, there is no economical way for the city to get a cheap, fast link to the Internet. Martin said Frontier owns the only fiber optic cables in that area, and Frontier does not make interconnection agreements — that is, it doesn’t lease unused “dark fibers” that are part of its private network to other entities — so the city of Philippi cannot reach a middle-mile bandwidth provider in a bigger city.
Philippi Communications System purchases its bandwidth from Citynet, but the data has to go through a costly 45 mbps link over Frontier’s network between Citynet’s fiber network and Philippi Communications System’s fiber network, which bottlenecks the speeds available to residents and businesses in Philippi.
Santa Monica, Calif. acquired a leased fiber interconnection with a major data hub in Los Angeles, allowing it to offer 100 mbps connections to businesses that once cost $3,500 a month through area ISPs for $500 per month through the city’s fiber network.
Philippi’s fiber network has no such link to the Internet backbone, leaving customers paying high prices for relatively low speeds that could be obtained more cheaply through Frontier’s DSL service.
Martin said much of rural West Virginia is only served by Frontier because without a link to open-access fiber, it is not economically feasible for new ISPs to enter the market.
“In their rural markets where they have no competition — their biggest audience base — why would Frontier spend an extra dime upgrading their service if they’re not going to get more money from that customer?” Martin said. “They have no reason to enhance their service or make it better. Introduce a competitor, and things change.”
Williams said high-quality broadband Internet access is a necessity for businesses of all sizes in today’s economy, and that he believes a lack of innovative service has stopped Huntington from reaching its full potential in recent years.
“We are full of businesses in our community that are world-class businesses that compete internationally,” Williams said. “Why should we expect anything less than that from the services that are provided to the public? That’s not acceptable. We should not settle for anything less than first-class, world-class.”
Williams said time will only tell if Huntington will receive that level of service, but that “we made it very clear what it is we expect.”
Comcast spokesman Adam Pratt was contacted for this story but declined comment. Minutes later, Pratt sent a statement to the Daily Mail indicating that Comcast would notify Huntington customers that it has increased Internet speeds at no cost.
Pratt said Huntington customers on Comcast’s “Performance” tier now have download speeds of up to 25 mpbs (formerly 15 mbps), customers on the “Blast” tier now have speeds of up to 105 mbps (formerly 50 mbps), and customers on the “Extreme 105” tier now have speeds of up to 150 mbps (formerly 105 mbps).
Contact writer Marcus Constantino at 304-348-1796 or firstname.lastname@example.org. Follow him at www.twitter.com/amtino.