Comcast Raises Invisible Data Cap

"Data Cap" - A Man Wearing a Cap Reading "DATA"

In what’s going to be seen as a response to strategic criticism, Comcast has raised the consumption cap on its residential broadband service plans from 250 GB per month to 300, the first increase since the cap was adopted in 2008. The increase is unlikely to be noticed by actual Comcast customers because the 250 GB cap wasn’t a problem. The increase hasn’t muted Netflix’ poorly-founded complaints of mistreatment.

Comcast has been criticized recently for its 250GB monthly limit on total downloads by residential broadband customers. Netflix complains that the limit is applied arbitrarily in order to keep it from growing. The Netflix complaints have been echoed by so-called public interest advocates Free Press and Public Knowledge, by the tech bloggers who generally take a hard line against networking companies such as GigaOm’s Stacey Higgenbotham, Ars Technica’s Timothy B. Lee and Nate Anderson, and the shy piracy advocates who post anonymously at TorrentFreak.

The Netflix argument is that 250 GB is so low that it discourages potential customers from using its service in favor of competing services such as the Comcast Video-on-Demand (VOD) service (operating through proprietary set top boxes, DVRs, and Xbox’s) that’s exempt from the cap because it doesn’t touch the Internet. The larger issue is whether managed services provided solely within the Comcast private network enjoy a  competitive advantage over such Over-The-Top (OTT) services as Netflix, Hulu, Amazon streaming, and Skype.

No one has produced a single consumer who has been been denied service by Comcast because of excessive use of Netflix, so the debate is at best theoretical and perceptual. In fact the FCC has not seen a single complaint that any networking company has violated its Open Internet regulations (despite blog gossip to the contrary.) It would be reasonable to conclude that Netflix is using the issue of data caps to obtain leverage over Comcast in order to force it to license additional content from its NBC Universal portfolio rather than raising an issue that has any real effect on its business. This is certainly an issue with a large empirical element: the current caps is either discouraging Comcast customers from using Netflix or it isn’t, and if it is there should be some actual evidence to prove it. We haven’t seen any, and as we read the complaints by the public interest groups and the bloggers, we notice that the arguments deal with psychological claims rather than measurable harms.

The Comcast byte cap is simple to the point of being crude, and its critics are right that it stayed at the same level for the first four years it was in effect, despite the dramatic (four times) increase in bandwidth that the conversion from DOCSIS 1.1 to DOCSIS 3 brought to the network during this time. If our expectation had been for a cap to be tightly tailored to conditions on the network with annual increases, we would certainly have been disappointed by this. Exceeding the cap for several months in a row also carries a severe price: the user can either upgrade to a business account for twice the price of a residential account or accept termination of service.

Whether these terms of service are reasonable depends on the number of people affected. For a long time, Comcast didn’t have a formal limit, but simply relied on a “you know abuse when you see it” method of identifying rogue customers who were streaming videos in both directions 24 hours a day. The Florida attorney general demanded an explicit limit, so Comcast picked a number that was far in excess of normal usage but still somewhat below the usage of users they considered abusive. Because 250 GB was so ridiculously high in 2008, there was no reason to implement a highly refined system that suspended it during off-peak hours or raised it every year. Comcast has a secondary system called “Fair Share” that uses Quality of Service adjustment for heavy users during peak periods that does the detailed work.

Annual increases wouldn’t make much sense in any event, because capacity upgrades in a DOCSIS network generally come about as new generations of technology are invented. These networks are very similar to mobile networks in this regard, and generation upgrades are typically five years in the making.

The most disturbing feature of the Comcast cap is the termination of service feature, but it’s only invoked if the user exceeds the cap for three months in a row (or more) and refuses to upgrade to a business class plan. The press and blog accounts rarely mention these features, as they’re not scary enough to drive traffic as well as a charges of “Internet death penalties” do.

So Comcast has raised the limit to 300 GB/month for its least expensive plans, and is trialing two alternate methods of dealing with users who exceed the cap:

  1. Switching them to higher speed plans with higher consumption caps; or
  2. Selling additional capacity at a flat rate per 50GB for every speed tier.

In markets where these options are not being tested (presumably for customer acceptance) the caps will be suspended altogether. No cap for me.

Predictably, Netflix is not impressed by cap-free service from Comcast outside the test markets:

A Netflix spokesman called Comcast’s policy change “a small step in the right direction,” but maintained that it does not solve the philosophical dispute.

“Unfortunately, Comcast continues to treat its own Internet-delivered video different under the cap than other Internet delivered video,” he wrote in an email. “We continue to stand by the principle that [Internet service providers] should treat all providers of video services equally.”

What will it take to make Netflix happy?

The Netflix reaction is frankly dishonest. There is no evidence, nor even a credible claim, that Comcast treats some Internet-delivered video services different from others. So Comcast already stands by the philosophical principle (as worthless as it may be) that all Internet video services are equal. The fact that a a private network uses IP doesn’t make it “the Internet” according to any regulatory definition the FCC or the Internet community has ever used. A TV program streaming from a Comcast server to an Xbox is not “Internet video,” it’s a managed service. The FCC knows this, the FTC knows it, and the Internet technical community knows it.

So Netflix is unlikely to obtain the lever they desire in their licensing program. But good try, folks, you eliminated the invisible cap on my Internet service for a while, not that it ever affected my behavior anyhow.

Print Friendly

About the author

Richard Bennett is an ITIF Senior Research Fellow specializing in broadband networking and Internet policy. He has a 30 year background in network engineering and standards. He was vice-chair of the IEEE 802.3 task group that devised the original Ethernet over Twisted Pair standard, and has contributed to Wi-Fi standards for fifteen years. He was active in OSI, the instigator of RFC 1001, and founder, along with Bob Metcalfe, of the Open Token Foundation, the first network industry alliance to operate an interoperability lab. He has worked for leading applied research labs, where portions of his work were underwritten by DARPA. Richard is also the inventor of four networking patents and a member of the BITAG Technical Working Group.