Let the Netflix debate continue! A report from network performance vendor Sandvine reveals that Netflix traffic accounts for 30% of all peak download Internet traffic in North America. That’s one source representing close to one third of all download traffic. Pretty amazing.

Sandvine’s “Global Internet Phenomena Report: Spring 2011” points out that ‘Real-Time Entertainment’ applications, represented most prominently by Netflix, now represent 50% of all peak download traffic, a 60% increase since 2009. The data highlights the dominant role entertainment now play’s on the Internet. If you believe Sandvine’s numbers, entertainment will be closer to 60% of all peak download Internet traffic by the end of this year.

These numbers help fuel the debate around issues like broadband metering and caps, net neutrality, and who ultimately pays for the massive amounts of bandwidth entertainment applications like Netflix now demand. Peer-to-peer file sharing runs a distant second to entertainment and social media is inconsequential, when put into the context of entertainment demand.

Broadband service providers are reacting to these trends in various ways, but instituting some form of broadband caps seems to be the order of the day. For smaller broadband carriers, the issue of entertainment downloading in general and Netflix in particular is of heightened awareness. Smaller regional and rural broadband carriers have limited backhaul capacity due to various reasons including distance, availability, and cost constraints. I’ve seen network monitoring statistics for rural broadband carriers that show applications like Netflix consuming well over 50% of total backhaul capacity – an alarming trend to some.

Some consumer advocates suggest the issue is overblown and broadband carriers are creating illusions of ‘scarcity’ to justify higher prices and better profits for broadband access. Like with any argument, there are elements of truth on all sides. Often times, all broadband carriers, regardless of size and scope, get lumped in together with this argument, which in my opinion is a mistake. Large national tier one carriers like Comcast and AT&T have very different exposure to this issue than a rural broadband carrier that has 2,000 total broadband subs with a 10 meg pipe to the Internet.

Whatever the argument, the implications caused by the growing demand of consumers to be entertained via the Internet are far and wide, with no logical conclusion in sight.



Join the Conversation

11 thoughts on “Report: Netflix Represents 30% of Peak Internet Traffic

  1. If I like to be "entertained" by having fresh, new water in my pool every week, and I replace the water in it quite frequently, doesn't my water bill go up accordingly? Just sayin'

    1. The economics are significantly different for broadband as for a physical good like water. Transmitting extra bits costs practically nothing whereas using more water does have an incremental cost. Those complaining about Netflix have ignored the need to upgrade their old copper networks and are now complaining that customers want to use what they have been promised for years.

      1. @ communitynets- The real issue, as I see it, is that networks must be built to accommodate peak load capacity, or the user experience suffers. Our average usage during the month might be "x", but we are multiples of "x" at peak times. If someone was watching Netflix mid day or at 1am, I would agree with you… the incremental costs are very low to deliver those bits to the home. But we must design and engineer the network to handle peak- and that is very, very expensive not only for internal network components, but especially good, quality Tier 1 bandwidth. And while the price per Megabit has gone down significantly, our bandwidth needs have risen much faster than the price decreases.

  2. Netflix making the most money? I highly doubt it. The profit margins of offering broadband over cable networks for a provider like Comcast undoubtedly exceed, by a wide margin, those of Netflix.

  3. Everything costs money and everything has capacity limits. Your "practically nothing" comment shows you have no idea of true costs. No matter what the transport, fiber, copper or the air, it all takes electronics and it all costs money to buy and maintain. Just as with the public telephone network, take the money out of it and no one will provide it. Now the government can step in and "save the day" using our tax money to waste capacity and efficiency. As with all things, the fairest way is for the user to pay.

  4. @communitynets That’s a fairly naive view of broadband. While the actual transport cost of the bits is fairly stable, the CAPEX associated with all of the electronics is not insignificant. The replacement/upgrade cycle of those electronics gets shorter and shorter thanks to apps like Netflix, resulting in quite the opposite of ‘practically nothing.’ Just go talk to some folks actually running a broadband network.

  5. "2000 subs with a 10 meg pipe to the Internet"? I only wish. How about 2000 subs with a 120 meg pipe to the Internet…………….driven by??????????????? Guess who?

    1. Yes, 2000 subs for 10 Mbps would be severely oversubscribed.

      Peak rates are currently 50 to 100 kbps/sub, so 2000 subs would suggest needing 100 to 200 Mbps. It was only two years ago where 25 kbps/sub was sufficient.

Leave a Reply

Your email address will not be published. Required fields are marked *

Don’t Miss Any of Our Content

What’s happening with broadband and why is it important? Find out by subscribing to Telecompetitor’s newsletter today.

You have Successfully Subscribed!