The Federal Communications Commission recently released a technical paper confirming what most technology insiders – and many movie fans who have tried to stream a high-definition Netflix trilogy at 8 o’clock at night — have long suspected. Actual download speeds on broadband connections in America lag way behind what most providers advertise.
On average, the FCC concluded, you’re probably only getting about half the speed you signed up for.
The figure highlights a unique quirk of the broadband industry in the U.S. General Mills isn’t allowed to sell you cereal labeled with “as many as” 5 grams of protein. Nor does Chase get to sell you a credit card with “as little as” 2 percent interest. So why do broadband providers get to peddle aspirational standards?
“You don’t go to a restaurant and order a meal and then only get half of it,” said Benjamin Lennett, a senior policy analyst with the New America Foundation’s Open Technology Initiative. “I can’t think of anything where you expect to get one thing, and you only get half of that. You don’t buy a sports car and then it only goes 30 miles an hour. And that’s what they’re selling here, because everything is about high speed, the fastest speed.”
Granted, broadband providers can’t promise you a fixed speed at all times; that number depends on how many of your neighbors are vying for a finite amount of capacity at any given time. Broadband providers also sell you an “up to” speed — say, 10 megabits per second — on the assumption absolutely every customer won’t really require that much all the time (or even any of the time).
Miller-McCune’s Washington correspondent Emily Badger follows the ideas informing, explaining and influencing government, from the local think tank circuit to academic research that shapes D.C. policy from afar.
The FCC calculated that 80 percent of broadband use never requires more than 4 megabits per second anyway — for now.
But broadband providers could exercise a little more truth-in-advertising by selling a guaranteed minimum speed, rather than a hopeful maximum one. That’s what they already do with businesses.
“Basically, they’ve set a floor for business,” Lennett said, “and then they set a ceiling for residences.”
The Open Technology Initiative has even designed a handy format that broadband providers could use to share with residential customers their newly honest advertising claims — a broadband nutrition label.
The label would standardize all kinds of information relevant to the savvy broadband shopper, including a rundown of all the installation, modem rental and early termination fees, as well as more sophisticated benchmarks such as “maximum round-trip latency.” But even the neophytes could compare advertised and minimum speeds.
Lennett envisions the required label appearing on provider websites, with your purchase agreement, even on your monthly statement.
The idea has been necessitated by the evolution of broadband technology — and the increasingly complex things users want to do with it.
“It’s not like the old telephone networks where it was based on circuits, each circuit got a certain amount of capacity, and once you ran out of circuits, you’d get a message saying ‘All circuits busy. We can’t connect your call,'” Lennett said. “That doesn’t really work on broadband or packet-switch networks because everyone shares capacity, and instead of getting ‘all circuits are busy’ messages, your speed slows down.”
Most users never notice this phenomenon as they’re browsing websites. But increasingly, that’s not all we’re doing on the Web. Today, we can stream live television, file share and play real-time video games with teenagers 12 time zones away.
As demand for speed to do all that increases, the FCC concludes that “consumers need a better, publicly agreed-upon measure of broadband performance” that reflects actual end-user experience. This is also a recommendation of the National Broadband Plan, and moving toward it the FCC is planning a more systematic analysis of actual speeds on individual computers.
For anyone who doesn’t want to wait on that report, you could also go here and try the experiment on your own.
