The average American's Internet service is about to get a whole lot faster. Digital content publishers such as Netflix (NFLX 4.17%) stand ready to reap the benefits of this upgrade, while telecoms from CenturyLink (LUMN) to Verizon (VZ 2.85%) are shaking in their boots.
FCC chairman Tom Wheeler wants faster broadband services. Source: FCC.
What's new?
Federal Communications Commission chairman Tom Wheeler is about to drop a bomb on the broadband service provider industry. In a proposal currently under review by Wheeler's fellow FCC commissioners, and presented to the tech gurus at Ars Technica, he wants to raise the bar for broadband services.
And it's not a small change.
This definition is supposed to be updated every four years and was last touched in 2010, when broadband services were defined as an "acceptable quality of service" with download speeds of at least 4 megabits per second, 1 megabit for uploads. It's now five years later. In other words, the FCC is overdue for a change.
The new limits? Wheeler is thinking at least 25 megabits download, paired with upload speeds of 3 megabits or more.
The idea is to keep Americans up to speed with the demands of modern Internet services. In 2010, Netflix had barely started slinging digital movies across the Web, and Google 's (GOOG 1.25%) (GOOGL 1.27%) YouTube served mostly standard-definition videos.
Today, it's all high-def, all the time, and online video is also much more popular than it used to be. As it happens, streaming video is a huge bandwidth drain.
If you're stuck at the minimum broadband level from the 2010 rule-making, you could easily suck up your entire bandwidth by viewing just one high-def video on a single screen. Watching two shows at once with acceptable quality becomes difficult, not to mention using the Internet for anything else while the YouTube or Netflix shows are playing.
So, Wheeler must update the FCC's broadband definition. It's not a purely academical change, either: Federal funds for rural broadband installations depend on this definition.
The difference between cable and phone networks
And that's where things start to get real. With money on the line, you can bet that the Internet service providers of America will sit up and take notice.
For the cable industry, that's not a very big deal. The infrastructure and technology behind traditional cable TV systems are easily convinced to provide 25 Mbps Internet speeds. These systems were designed to support dozens of bandwidth-hungry analog TV channels, which translates into extremely high connection speeds in the digital world. Your local cable provider can probably sell you an Internet package with double or quadruple the proposed new minimum speed.
Telecoms, on the other hand, will be left between a rock and a hard place.
The DSL technology behind most telephone-company Internet services was never meant for ultra-fast connections. Designed to work with twisted-pair copper phone lines, this technology runs into a brick wall very quickly. And the further away you live from the phone company's local connection center, the lower your network speeds will be.
The cheap and easy way to set this up is to attach a DSL access multiplexer, or DSLAM, to old copper wires that were installed decades ago with no high-speed Internet services in mind. Unless you happen to live within a couple of miles from the DSLAM, you'll barely qualify for the old definition of broadband services.
Twenty-five megabits can be done over DSL lines, but only when the DSLAM sits right in your neighborhood and preferably just a couple of blocks away. Researchers and hardware developers have shown gigabit speeds over DSL, but only under laboratory conditions and at very short distances.
If you live in a Verizon service area, but not in a neighborhood blessed with FiOS fiber-optic services, Big Red simply won't sell you a 25 megabit plan today. Its fastest DSL service stops at 15 megabits down and 1 megabit up, and even that underwhelming option is "available in select locations only."
Rural and regional telecom specialists are no better off. For example, CenturyLink's DSL speeds often stop at 20 Mbps. According to a recent conference presentation by CenturyLink CFO Stewart Ewing, only 40% of his customers can get the 20 megabit service.
Verizon CFO Fran Shammo crystallized the situation in another recent conference show. "Outside of FiOS, where I only have copper to compete against cable, I'm not going to win that battle," Shammo said. "I can't compete on speed. We've made the strategic decision not to invest in that copper plan anymore."
Fiber, fiber, fiber!
Ultimately, telecoms will need to adapt. At the very least, that means pulling high-speed optical fiber connections into every neighborhood, stick a DSLAM at the end of that line, and rewire the neighborhood to route the local copper cables through this connection center.
But doing just the bare minimum today will lead to another headache four years down the road, when the FCC updates its broadband rules again. By then, digital 4K video should be commonplace and placing much higher demands on average connection speeds. Barring serious improvements to the basic DSL technology, even a local DSLAM just won't work anymore.
So the telecoms might as well pull fiber straight to your curb in the first place. Verizon's FiOS is available in speeds up to 500 megabits per second. CenturyLink's fastest fiber-based plan runs at a full gigabit. In the long run, fiber-optic services like FiOS and CenturyLink's gigabit plan are the future for telecoms with broadband ambitions. This way, it'll be many years before the infrastructure becomes outdated again, requiring another set of expensive upgrades.
But therein lies the rub: Fiber pulls are expensive.
Verizon's FiOS service is available to nearly 20 million American homes. That's very close to filling out Verizon's total market footprint of about 21 million homes, and just about as far as Verizon is willing to take its FiOS investment. The company claims to have spent more than $23 billion on its FiOS installations.
So Verizon is on the right track. The company has either converted or sold off its legacy copper systems over the years, leaving it with a mostly fiber-based infrastructure. Big Red's aging DSL lines are simply going away. Twenty-five-megabit services are easy to come by when you're sucking down fiber-optic light pulses. So Verizon shouldn't fear Wheeler's broadband update too much, since the new policy basically underscores what Verizon has been doing for the last few years anyhow.
Smaller telecoms with a regional or (shudder) rural business focus may not be so lucky.
Take CenturyLink. The company offers a TV subscription service wherever it lays down fiber-optic broadband lines. That TV product is currently available to about 2 million homes, with about 10% of those households actually paying for the service. In other words, roughly 3% of CenturyLink's 6.1 million broadband customers are using its fiber product today.
The other 5.9 million broadband users will largely see their access slipping below the new broadband classification. So to sell its Internet services under the "broadband" banner, or to qualify for federal infrastructure funding, that tiny fiber market must expand dramatically.
CenturyLink has been upgrading its infrastructure with fiber-to-the-node (think "fiber to the neighborhood") technologies in recent years. This project doesn't come cheap, as you can see from CenturyLink's capital expense trends:
CTL Capital Expenditures (TTM) data by YCharts
What's next?
If Wheeler pulls the trigger on this huge broadband definition update, it's good news for consumers who can expect better service but bad news for cash-strapped telecoms.
And even the cable guys and fiber-focused telecoms will find it harder to justify high prices for what used to be high-end services. As the lower end of the broadband market rises to meet the new requirements, the top shelf must also move higher up.
Looking ahead, there's still much to be done before you can call the American broadband market competitive or cost-effective. From an international point of view, America's high-speed Internet access is kind of laughable. Just ask Tom Wheeler:
"There is an inverse relationship between competition and the kind of broadband performance that consumers are increasingly demanding. This is not tolerable," he said at a presentation last September. "At 25 Mbps, there is simply no competitive choice for most Americans. Stop and let that sink in ... three-quarters of American homes have no competitive choice for the essential infrastructure for 21st century economics and democracy. Included in that is almost 20% who have no service at all!"
So Wheeler is doing what he can to correct that problem. Telecom investors had better pay attention, because what's going on here will not be good for the current business model with high profit margins and limited competition.
