Is your ISP cheating you out of bandwidth? New gov't test offers a hint

When you hear a politician, a lobbying group, or a corporate executive yammering on about leaving the free market alone, don’t fall for it. Many businesses today are built around finding that industry -- or that particular anomaly within an industry -- where market forces simply don't apply. Then, they exploit the heck out of it for big profits. Late fees on credit cards, for example -- who shops around for those? Or hotel safe fees that get automatically tacked onto your bill. Call this the "un-free market" at work.

Broadband Internet access has long existed in this un-free market. For a long time, many U.S. consumers had only one Internet access provider, the very definition of an un-free market. Now, most people have two options -- an improvement, but duopolies rarely behave like free markets.

More important, when people shop for broadband, they usually don't know what they are getting. What does 10 Mbps speed really get you anyway? Is it five times better than 2 Mbps? Should you pay five times more for it? And most important, since the ISPs guarantee none of this, what do consumers really get for their $59.99 per month?

Imagine pulling up to a gas station and paying by time rather than volume. A 60-second shot of gas costs $20 -- sometime you get 5 gallons, and sometimes you get 3 gallons, but there no scale telling you how much you get. That's pretty much how things work in the Internet access world right now.

Finally, government regulators have stepped in and begun making the purchase of Internet access fairer. The Federal Communications Commission released the results of a year-long, scientific study of 13 U.S. broadband providers, focused on whether firms like AT&T, Verizon, Qwest, Frontier, and others live up to their marketing materials when delivering high-speed Internet service. For the first time, the firms are being held to their promises, and consumers can find out if they are getting what they are paying for.

The results weren't bad. Most firms hit 90 percent of their advertised upload speeds. Download speeds were another story, however. Only four of the 13 reached their promised download speed (Charter, Comcast, Cox, and Verizon), while most of the other firms hovered around 80 percent. Cablevision fared worst, with download speeds that fell to 50 percent of advertised rates at peak periods. ( is a joint venture with Microsoft and NBC Universal, the latter of which is a division of Comcast.)

Overall, the tests also showed that consumers aren't getting dramatically cheated by their ISPs, and time-of-day tests showed there is no longer a massive web traffic jam every evening when consumers get home and log in. The tests showed that once-dreaded peak period slowdowns are now uncommon.

Still, Free Press research director S. Derek Turner, who had long called for such government tests, was critical of the findings. Free Press is a nonprofit group that advocates for consumers on a number of media and telecommunications issues.

"The way we look at the results, most of the ISPs -- nine of the 13 -- are failing to deliver their promised speeds at all times of the day. The fact that four of the 13 can deliver means the others could be," he said. "If they are unable deliver on their promises, than their advertising is highly misleading. ... If you got 80 percent of your pay for work you did, you wouldn't be happy about it." 

The tests, which were performed by a private firm that has run similar tests in the U.K., measured performance at 6,800 representative homes around the country in March. The tests revealed a clear trend: DSL services were least likely to live up to advertised rates. Fiber-optic services, like Verizon's FiOS, fared best, followed by cable-based systems. Wireless broadband services, like Verizon's Mobile Broadband, were not tested.

Outside observers said the test results seemed consistent with real-world experience. Damon Mueller, who operates TestMy.Net service, says he thought the FCC used sound methodologies. TestMy.Net gives users free, instant feedback on their Internet connection speed, and lets consumers compare results with others.

"The FCC's test is accurate and their results mirror my own. They even mention the ISPs that top TestMy's list," he said.

There are severe limitations to the FCC's results however. To borrow a phrase from politics, all ISP issues are local. An ISP might offer great service in New York City, but lousy service on your block in Queens because of dated wires on the telephone pole. In other words, it's quite possible that DSL, while performing worst in this test, could be the best option at your house.

Consumers, as always, bear some responsibility, and the FCC's research shows they are dangerously detached from their broadband purchasing decision. Fully 80 percent said they did not know what speed they purchased from their ISP; another 49 percent, when asked to identify their level of service, got it wrong.

This ignorance can be overlooked, however, given that the speed numbers have proven to be rough approximations of the service provided. And besides, only geeks really care about the speed of Web traffic flowing in and out of their homes. Most consumers simply want to know that they have enough reliable bandwidth to do what they need to do.

"To most consumers, megabits are a very abstract concept. They just want to know, 'Does it work?' Can I use Netflix?' " Turner said. 

To that end, Mueller is worried that the FCC results -- while a good test to make sure the ISPs aren't committing fraud -- aren't really that helpful to consumers. 

"The data still isn't being presented in a way that helps people make the best choice in service provider," he said.

Here's the kind of bizarre calculus consumers are still forced to perform: Paying $60 for a service that only achieves 50 percent of advertised rates -- say, a 10 Mbps fiber optic connection -- could still be a better deal than paying $50 for a 2 Mbps service that hits 100 percent.

The FCC's download test results.

And again, none of that information actually promises an uninterrupted Netflix movie. To really create a free market where consumers could intelligently shop for services in their local areas, free trial periods that allow side-by-side, real-world comparisons would be necessary.

It should be clear that long-term commitments, like two-year contracts required by some Internet providers, are the death of such a free market.

That's why, to Turner, the FCC tests are only the beginning. 

"We would like the FCC to pressure underperforming companies to explain why they didn't do well, and to disclose that the upper limits they advertise are never attainable," he said.

One hopeful sign that the test is already setting free market forces free -- two days after the results were published, Verizon sent e-mails to some customers bragging about the results, Turner said.

"And, hopefully this will embarrass (other ISPs) into performing better next time," he said.


It's a good idea -- not to mention fun -- to test your broadband connection every month or so, to make sure you are getting what you pay for. There are several services available, like Mueller's TestMy.Net. Along with offering useful points of comparison, the site can also store results for comparison later. The Speakeasy.nettest, now operated by ISP MegaPath, is among the most popular and easy to use. And the FCC now offers its own broadband test.

Even if you happy with your current service, you should still perform the test. Who knows what great, bandwidth-hogging service might catch your fancy next month?

What should you do if find your ISP is underperforming? Ideally, you would switch to a better service, but there's no guarantee that will improve your situation. A competitor could be exaggerating speeds, too

You can always complain that your service is too slow, but know that it's standard industry practice for ISPs to use such calls to “upsell” customers on more-expensive plans.

"Those types of stories are pervasive," Turner said. "Customers call to complain that something they are trying to run isn't working, and customer service agents are trained to upsell them."

Be ready for the sales pitch. First insist that the company live up to its promises. Make it test your service, and compare the tests with your own results. In the end, an upsell deal might be worth taking, but be a good “free marketeer” and avoid the long-term contract.

Follow Bob Sullivan on Facebook or Twitter.