Is Your Internet Connection Slower Than Advertised?

Categories: Science

It's a common occurrence these days: you sign up for a specific level of service with your local internet service provider (ISP) and expect to get the speeds advertised. But most of the time, you don't.

On Monday, the Wall Street Journal published a report that takes that hassle from individual anecdote to evidence, with a lengthy test of internet speeds across the country: for most ISPs, the result wasn't pretty.

According to the Wall Street Journal's extensive survey from Ookla — better known as Speedtest.net to the frustrated internet speed-conscious out there — which "compiled data form tens of millions of speed tests as well as surveys of 646,404 Speedtest users," most internet service providers' real service just doesn't match up.  "Indeed," says the report, "the vast majority of the 800 cities...experience median Internet speeds that are slower than what their providers advertise."

Out of the 800 U.S. cities and 27 ISPs the study took a look at, only seven reached or exceeded the internet speeds advertised, on average.  That means 20 ISPs were not up to snuff, with some major offenders on the list (take a look at WSJ's report for an incredibly detailed listing, city-by-city, of actual speeds versus advertised speeds).

The worst offenders?  Clear Wireless, a Sprint-owned 4G wireless company for home internet devices, maxed out at an average of 41 percent below the advertised speed.  Next were Windstream Communications (another wireless-emphasizing broadband company) and Cable One, which provides broadband in 19 states.  Verizon and AT&T Internet Services (DSL) were the two major internet providers also at the bottom of the slow list.  Time Warner Cable and Comcast — two of the biggest cable internet providers who have proposed a merger — were on the negative side, but just barely.

(Ookla, via WSJ)

The most consistent company was cable company Cox Communications (also named recently by Netflix as one of the top average ISPs for its streaming service), and at the top of the list, Midcontinent Communications, serving Minnesota and the Dakotas, which delivered average speeds at least 8 percent or faster than promised.  Perhaps Midcontinent will want to adjust its marketing strategy and stop selling itself short.

Caveats

More than half a million people surveyed and tens of millions of speed tests make up a pretty large sample size for a study, but there are some major caveats to consider that may poke some holes in the study's veracity, or at least be a cause for a little grain of salt.

When Do You Test?

I don't know about most people, but I usually run a speed test when my internet seems to be running — you guessed it — slower than advertised.

Equipment Issues

A lot of the time, the slow speeds are due to a modem that needs to be reset.  In addition, running a Flash-based site like Speedtest.net can also be taxing to older computers that can't run the software effectively.  Slow, old, or improperly configured WiFi routers can mess up an otherwise good connection, too.

Server Selection Issues

Speedtest.net will automatically choose a server to connect to for its test, if you don't choose one yourself, but sometimes the site's software will astonishingly choose a server that's farther away or simply slower than one you could pick yourself.

Data Caps

Whether this is fair or not is for a different discussion, but your ISP may have a monthly data cap.  If you exceed that cap (perhaps by torrenting too much), the company will throttle your connection: As advertised.  That fact may be in the fine print, but it's there.

How does your ISP stack up to its own hype?  How does it stack up to the other ISPs in the study?  Let us know in the comments!


User Comments:
Jonas - Oct 10, 2014 12:25:48 — Despite Cox Communication's seemingly perfect score, they failed (again) in my experience. I had subscribed to high-speed Internet access in March 2014. Up to 150Mbps, they advertised. But the speeds tests I ran regularly averaged around 40Mbps, and as low as 3.53 Mbps. Despite several calls to technical support, a new modem, a new router, and new cables, Cox could not get the speeds up to where they promised. Cox claimed that my tests, which were run through XFinity, Speakeasy, and Ookla (the same speed test Cox uses), were invalid because they were not limited to Cox's network. Why should I pay to get 150Mbps on Cox's network, but only 40Mbps on the rest of the Internet? So I fired Cox and I'm happy to say that I'm now Cox free. I dream of the day when we can have a free market and true ISP competition in America. Someday...maybe.

© 2010-2013 Jonas Maxwell. All rights reserved.