The State Of Broadband update from July 2010

We all remember the bad old days of having to load data into removable media in order to send it off to the data center. After all, it would have taken days to transmit the necessary data over slow telecom links.

Problem is, the bad old days aren't over. Instead of shipping tapes to data centers, organizations now regularly ship entire hard drives to cloud providers. Despite tremendous advances in line speeds, it still can take a week or more to transmit very large data sets, even if your line speed is 10 Mbps. Of course, companies don't regularly need to transfer terabytes of data over the internet, but the current level of sneakernet that's necessary for the transfer of even a few hundred gigabytes seems a bit high for the 21st century.

The state of broadband matters to your organization. There's been considerable consumer interest over the past several years, culminating in an FCC plan announced earlier this year to expand broadband coverage and speeds and promote competition. IT organizations can benefit by staying in touch with those regulatory issues, as well as taking advantage of new technology trends, such as wireless broadband, and partnering with alternative providers and municipal networks that buck the status quo. There are clearly risks in doing so, but taking no action almost guarantees that enterprise IT, with pockets of presence in rural and other nonurban areas, will continue to be held back by low-capacity, high-expense networks.

There are many reasons why the state of consumer broadband should matter to enterprise customers:

>> Companies need better broadband into their offices, factories, and warehouses, but they should also worry about the quality and cost of broadband that consumers get, for several reasons. First, their employees are using it, either as true telecommuters or for that bit-of-work-in-the-evening, especially across time zones. Second, excellent consumer broadband will change how your company can interact with its customers and the services it can offer them. For example, telemedicine is a big potential application; a few insurance companies are starting to reimburse for online "house calls," but spotty video connections won't allay patient fears that they're getting second-class care.

>> Besides telemedicine, applications like video streaming, off-site backup, videoconferencing, and remote education will greatly expand bandwidth needs--not only at corporations, but also for the small office and home office (SOHO) market.

>> IT has a vested interest in how much connectivity costs, and the level of competition depends significantly on how broadband regulations play out.

>> While the consumer model creates more competition and lower pricing, it doesn't come with the assurances of a service-level agreement that companies need. Enterprises that dabble with SOHO lines should be paying attention to the current brouhaha over whether carriers can prioritize traffic according to their own management policies, including which customers pay the most. As we write this, the FCC is trying to reassert its authority to regulate whether a carrier can prioritize traffic. The results will be important to enterprise customers. Keeping an eye on how providers are allowed to apply usage caps will be important, too.

>> The American Recovery and Reinvestment Act's BTOP (Broadband Technology Opportunity Program) and BIP (Broadband Initiatives Program) matter to large enterprise in ways that you wouldn't consider. Read on.

Problem With The Status Quo

In June, National Cable and Telecommunications Association CEO Kyle McSlarrow called America's broadband deployment over the last 10 years "an unparalleled success story," alluding to the rise of cable IP networks and faster and more extensive broadband in the consumer market. He's right by some measures. Among the G7 countries, even though the U.S. is only No. 5 in broadband penetration (see chart on previous page), it's been making headway. But when you look at average broadband prices worldwide, the U.S. doesn't compare favorably--service in the United Kingdom, Sweden, France, Japan, Korea, Germany, and many other industrialized countries is cheaper, on average. And when you look at broadband subscribers per 100 inhabitants, the U.S. is ranked No. 22, slightly above the Organisation for Economic Co-operation and Development average but below the Scandinavian countries, Korea, Canada, France, the U.K., and others.

As with many things, where you stand depends upon where you sit. Tony Patti, CIO for S. Walter Packaging, a century-old manufacturing company in Philadelphia, says that even in the SOHO market, significant bandwidth is for sale relatively cheaply (see chart, below). "People always want more for less, but we're at a remarkable and revolutionary time in the history of the convergence of computing and communications," Patti says. But the two key questions are these: Are you in the provider's service area; and if you are, does the actual speed match the advertised speed? In major markets, the answer is: probably. But talk to someone in smaller cities and rural America, and a different story emerges.

Kris Hoce, CEO of Pardee Hospital, a 200-bed facility in Hendersonville, N.C., says the hospital's telecom lines are "stretched" today, and when the management team looks at tomorrow's challenges, including telemedicine and telemetry, he gets even more concerned.

Until a second competitor, Morris Broadband, entered the market a year ago, the incumbent provider was Pardee's only option. "You'll take whatever capacity they give you, do it on their time schedule, and you'll pay through the nose for it," Hoce says. Since Morris Broadband's entry, Pardee has realized a 10% to 15% reduction in telecom costs, though it can't always get sufficient bandwidth, he says.

It's hard for competitors to break in. We spoke with several small telecom providers that had problems getting what's called "egress" from the incumbent provider's point of presence in an area. The problem wasn't getting a circuit to the POP, or building infrastructure to serve a customer a mile or two away. The problem was getting the incumbent to let the competitor run fiber into the building. The incumbent owns the POP and can say what can or cannot be connected to it. The U.S. doesn't have "unbundling" regulations (broadband slang for rules that force providers to choose between being providers of the "highways" versus the "off-ramps") the way that other countries do.

Some large enterprises make the classic build-versus-buy assessment--compare the cost of leasing circuits (high) with the cost of obtaining a fiber IRU (indefeasible right of use, essentially permanent ownership). Then if the incumbent carrier refuses to sell them an IRU, they compare the cost of the leased circuits over time with the cost of putting in their own fiber.

That scenario played out for North Carolina Research and Education Network, a not-for-profit that operates statewide. Leased circuits would have cost NCREN $85.4 million; to build, operate, and maintain that same fiber cost $39.9 million, says Tommy Jacobson, NCREN's chief architect. It's easy to see why it chose to build.

Of course, that scenario isn't realistic for most smaller organizations, or for organizations with a larger footprint than one state. But it does show how the economies of connectivity are out of whack--it shouldn't be cheaper to build out fiber than to buy the bandwidth, even over a long payback period.

National Broadband Plan

The FCC's 376-page National Broadband Plan, while a testament to the ability of federal bureaucracy to fill large amounts of paper, stands to benefit enterprise IT over the next few years in several areas, if the agency follows through.

First, the FCC says that it will be publishing market information on broadband pricing and competition. Will this be as useful as PriceWatch and eBay are in determining what you should pay? We're not sure. But transparency itself should help: A market where all players know what everybody's charging tends to be one where prices dip as low as possible.

Second, the FCC says it will make additional wireless spectrum available, and it will update its rules for backhaul spectrum. President Obama has thrown his weight behind this movement, directing the National Telecommunications and Information Administration--the folks behind the broadband stimulus--to help the FCC with a plan to make 500 MHz of spectrum available by the fourth quarter of this year.

It's unclear what the licensing procedures will be, and for which portion of the additional spectrum. Our bet: some mix of unlicensed spectrum (like 2.4 GHz, a nightmare for IT departments that want to avoid interference), some fully licensed (like 800 MHz, whose paperwork can take months or years to get processed), and some "lightly licensed" (like the 3,650-MHz band that was allocated for WiMax in 2005, which requires two or more licensees in the same region to cooperate). When additional spectrum comes online, it should revitalize the market and create product innovations, which should make broadband wireless a bit less difficult for enterprises to deploy.

The FCC also plans to improve rights-of-way procedures. Power and other companies that own poles either have undocumented or onerous agreements for anyone wanting to attach to a pole or bridge. Streamlining and standardizing this process would be welcome news to telecom market entrants and user organizations that want to bypass the telecom providers. The unanswered question is, how will the FCC "encourage" rights-of-way owners to improve these procedures?

The National Broadband Plan also stipulates longer-term (within the next decade) goals, including that 100 million consumers are able to access affordable 100-Mbps actual download speeds, 50-Mbps upload--more than 10 times faster than what most U.S. consumers can now get. More interesting to enterprise IT, the plan outlines a goal of affordable access to 1-Gbps links for "anchor institutions"--hospitals, community centers, schools, and so on. As these institutions get affordable links, other large institutions, like big companies, will also get affordable high-speed links.

The FCC doesn't always have the authority to say how these goals will be accomplished. But in the "implementation" chapter of the National Broadband Plan, it suggests who (including the FCC) should pursue them. For example, it recommends that the executive branch create a "broadband strategy council" consisting of advisers from the White House and its Office of Management and Budget, NTIA, FCC, and other agencies. The FCC also has committed to publishing an evaluation of its progress as part of its annual 706 report, named after section 706 of the Telecommunications Act of 1996. You can track 706 reports at www.fcc.gov/broadband/706.html.

Emerging Competition

Simplifying and streamlining the status quo won't be as quick as we want it to be, but the situation isn't bleak.

True, many of the wireline highways are owned by the same folks that own the off-ramps and have a big interest in resisting competition (the likes of AT&T, Verizon, and Qwest from the telco sector and Comcast, Time-Warner, and Cablevision from cable TV). But competition is in fact emerging.

Players like Morris Broadband serve relatively small and rural areas, catering to customers the larger players simply won't touch. CenturyLink, a larger player, serves rural customers in 33 states. PAETEC competes in 84 of the top 100 areas, known as "metropolitan service areas," which are anything but rural. Then there are municipal broadband projects such as LUS Fiber, a fiber-to-the home network started by the utility in Lafayette, La., that offers business services (10-Mbps symmetric) starting at $65 a month.

It's hard to get information out of the incumbents--we tried, but folks like Verizon said that they don't see how consumer broadband is related to serving enterprise customers. But the conventional wisdom is that they won't serve an area unless they can get 25 potential customers per mile. Smaller players will look at areas with five or 10 potential customers per mile. Bottom line: Whenever competitors enter a market, prices fall. In a striking irony, the incumbents opposed to broadband regulation have lobbied local and state authorities to prevent broadband buildouts by municipal entities.

In addition to the wireline broadband alternatives, consider that the airwaves are wide open. Wireless ISPs like Clear and mobile phone and 3G data providers like T-Mobile and Verizon Wireless are interesting, but your bandwidth and reliability may vary when attempting to use their business-class SOHO service. That said, back in the day of the bag phone, nobody would rely on a cell phone for anything that was hugely important, but that didn't keep IT organizations from playing with them in noncritical areas.

We're also interested by the services offered by the likes of Texas-based ERF Wireless, which is completely focused on serving businesses, mainly banking and oil companies. ERF's model: Customers invest in their own wireless infrastructure to backhaul to ERF's network and then pay an ongoing port fee to access a secured backbone. CEO Dean Cubley says ERF's banking customers pay about half of what they were paying to incumbent providers and have about a three-year payback on their capital investment.

The point is, it's not just about regulation, the FCC, and the incumbents. The best way forward might be to rely on alternative wireline and wireless carriers to increase broadband competition.

Economic Development Effect

The two federal broadband stimulus programs, BTOP and BIP, will make a substantial difference in our three criteria of competition, accessibility, and fairness. Most notably, the awards include provisions that any backbone network buildout must allow open access.

The programs will affect your company, if not immediately, then through what we think of as a "trickle-up" effect. Here are a few examples.

The Keystone Initiative for Network Based Education and Research, a Pennsylvania recipient of a $99.6 million BTOP grant, says it will serve "educational, research, healthcare, and economic development partners." These types of organizations used to talk only about education and research. Then, because of affiliations with university hospitals, healthcare got added. Hello, telemedicine! Now economic development organizations are getting involved. Thus, a trickle-up effect.

Even when stimulus dollars aren't in the picture, economic development folks get charged up about high-speed access. Electric Power Board, the municipally owned utility in Chattanooga, Tenn., says it will offer a fiber network with speeds of 150 Mbps to all 170,000 homes and businesses (urban and rural) in its footprint by the end of this year.

J.Ed. Marston, a VP with Chattanooga's Chamber of Commerce, says the utility's fiber network is integrated with a smart grid, letting it provide affordable electric power redundancy, which is "getting a lot of attention from both commercial and industrial users and has been very well received by economic development prospects." One ISP we spoke with who's familiar with the matter says Volkswagen chose to build a plant in Chattanooga (construction started in 2008, with $1 billion of investment and 2,000 new jobs) in part because of proximity to the fiber backbone that goes from Atlanta to Chicago, both significant network access points. Existing midsize (100-plus employees) and large businesses in Chattanooga are now more interested in developing work-from-home opportunities for their employees.

Jacobson of North Carolina not-for-profit NCREN says the group's successful BTOP round 1 application (awarded $28.2 million) came from efforts by the state's office of economic recovery. It's going to trickle up to the hospitals, too. "All the medical schools in the state are on NCREN today," he says, and "the nonprofit hospitals will be eligible to interconnect to us as well."

Welcome Back To Sneakerville

Some caution is necessary. There will be no shortage of poorly conceived broadband initiatives. Savvy IT organizations will stay close to operations, leaving the speculation to investors and economic development types.

Moving beyond sneakernet will require more than just fatter pipes. "Civil engineers discovered some time ago that building more lanes on highways does not really relieve traffic problems," says Mark Butler, director of product marketing with Internet services company Internap. "Relief comes when you use the available capacity in a more efficient manner."

So as you keep track of the legislation and other craziness coming out of Washington, keep pace with technical realities, lest you invest in higher-speed lines only to find that your use case isn't quite as you had planned. George Bonser, a network operator with mobile messaging provider Seven, cites cases of companies that install high-speed lines and then discover they can't get anywhere near their theoretical limit because of the software in use. It's a complicated matter that deserves your attention in the same way that keeping track of broadband competition, accessibility, and fairness does.

The (Very Messy) State of Broadband

Get This And All Our Reports

Become an InformationWeek Analytics subscriber: $99 per person per month, multiseat discounts available.Subscribe and get our full report on the state of broadband.

This report includes 19 pages of action-oriented analysis and charts, including:

Get This And All Our Reports