38 Comments

Summary:

Video is driving the projected increase in both mobile and wired broadband, but it’s not only the proliferation of video that’s the problem for mobile operators, it’s the relative ease that consumers now have accessing it. And that’s causing mobile operators to rethink their pricing plans.

Video is driving the projected increase in both mobile and wired broadband — but it’s not the proliferation of video that’s the problem for mobile operators so much as the relative ease with which consumers can now access it. Indeed, while mobile operators have long faced traffic congestion at cell sites thanks to peer-to-peer traffic, the widespread availability of video in formats that the average consumer can watch has changed the industry. And that’s causing mobile operators to rethink their pricing plans (GigaOM Pro, sub. req’d). In short, YouTube may be the death of unlimited mobile broadband on handsets.

Mobile video streaming rose 99 percent between the first and second half of 2009, according to data released this week by Allot Communications. The firm, which sells network management gear to broadband providers, credits the accessibility of YouTube and its ilk for the rise in video streaming on mobile networks. Video overall comprises most of the mobile network traffic, but the amount consumed via peer-to-peer traffic has fallen as the amount of streaming video traffic has risen. Jonathon Gordon, vice president of marketing with Allot, says that P2P has decreased as a percentage of mobile network traffic although it still comprises 19 percent of total traffic.

The reason for P2P’s gradual decline is that it’s more complicated to share files via P2P, which somewhat limits the audiences that will practice file sharing. For example, it requires finding and downloading software to a mobile phone, which not everyone is willing to do. YouTube, however, can be accessed by anyone in just a few clicks. As such, YouTube traffic accounts for 10 percent of all the traffic on mobile broadband networks, and 32 percent of all HTTP streaming traffic. And it rose 90 percent between the first half and second half of 2009.

Allot’s data proves that YouTube is a force today, but the latest numbers out from Cisco’s Visual Networking Index show the effects of streaming video on mobile broadband networks through 2014. And those effects are pretty brutal. Cisco estimates that 82 percent of mobile broadband traffic will be HTTP streaming traffic while total video traffic will account for about 2.3 exabytes of data a month.

Streaming traffic is more difficult for operators to manage simply because, as opposed to a video download, streaming is also an ongoing process. For ways companies are trying to improve this process check out our report on Adaptive Bit Rate Streaming (GigaOM Pro). Such real-time consumption of video during streaming has big implications for mobile operators’ networks, notably in that it can cause problems during periods of the day when other people want to use the same mobile network to surf the web, make phone calls or check email.

It’s no accident that AT&T’s Ralph de la Vega singled out video streaming during a speech his answer to a question asked at an analyst event, in which he attacked the gratuitous use of network resources by iPhone owners last December. In his speech response, de la Vega said:

“I’m not going to give you in detail what we’re going to do, but if three are causing 40 percent, then we’re going to try to focus on making sure we give incentives to those small percentages to either reduce or modify their usage so they don’t crowd out the other users in those same cell sites,” de la Vega said. “You’ll see us address that more in detail in the future…What’s driving usage on the network and driving these high-usage situations are things like video or audio that keeps playing around the clock. We’ve got to get to those customers and have them recognize and change their patterns.”

I’ve suggested that AT&T might use pricing as a means to shape user behavior on the network, rather than simply forbid users from doing what they want on mobile phones. Indeed, AT&T (and other carriers) may find itself racing to keep margins high for mobile broadband as usage increases. On its fourth-quarter earnings call at the end of January, Ma Bell admitted that its data traffic had doubled while the costs to send bits had halved. So for now, AT&T is keeping its costs in line with demand. But according to a forecast from Cisco released today, the average amount of data consumed on mobile devices will rise to 7 GB per month by 2014 from just 1.3 GB per month today — a 438 percent increase. Can AT&T — or other operators — drive the cost of bits down in line with that amount?

Given that mobile resources are constrained by a variety of things, including the spectrum allotted to carriers, it’s likely that mobile broadband providers will eliminate flat-rate pricing for mobile broadband as away to keep profits and network quality up while data use expands. When that happens should we blame YouTube — or profiteering mobile operators?

Related GigaOM Pro content:

  1. “But according to a forecast from Cisco released today, the average amount of data consumed on mobile devices will rise to 7 GB per month by 2014 from just 1.3 GB per month today — a 438 percent increase. Can AT&T — or other operators — drive the cost of bits down in line with that amount?”

    Given that Cisco is in the business of selling gear to move around bits, you think that forecast might have attempted to forecast the answer your question.

    Or maybe they, like AT&T are totally jazzed that articles like this push the “inevitability” of metering, when the evidence for it is paper thin. Given that mobile date increased from its early days a bazillion percent (an increase from zero = infinity), and costs have declined precipitously, I think they will be able to keep up.

    And if the market truly is as competitive as these carriers claim it is, then these absurdly high margins are bound to come down as the market matures. And if it is competitive, then investment will flow.

    Share
  2. I guess all we need to do is buy millions of $$$ worth of hardware and software from Cisco and Allot Communications to bail us out of this exaflood.

    I really would have liked to see some truly independent analysis and data in this article rather than just marketing material from those with vested interests.

    Share
    1. Oh and the Cisco ad at the bottom of these comments doesn’t lend much to your objectivity.

      Share
  3. Well, in my mind there’s no doubt that the first service provider to reintroduce variable mobile data pricing plans will lose 10% market share there and then. Not sure anyone can afford that.

    I am surprised however that they’re not looking into tiered pricing schemes more seriously. The advantage of tiered pricing is that the certainty remains, but different customers can be prioritized in different ways.

    Then again considering some recent bills of upwards of 140k EUR/month for a residential customer of a major European incumbent recently, I tend to think that their marketing teams need some serious brainpower injected into their pricing committees…

    Share
  4. Or we could just solve the problem using Photons instead of Electrons with Lightwave Logics (LWLG) Technology. Remember those words as you will be hearing them in the very near future.

    Share
  5. Starting to stop flatrate plans is definitely the wrong approach. What we have learned for example in Japan that exactly these flatrate plans on 3G networks were the the drivers for the growth of the mobile internet. Now discussing putting an end to flatrates in the U.S. or Europe would kill the mobile revolution before it even really started.

    First of all make sure your mobile ecosystem is developing instead of trying to squeeze money from mobile web consumer just after things started to pick up a bit.

    Share
  6. I agree with Stacy – I guess service providers will not have a choice but charging the customers for data services in order to fund the massive investments in hardware.
    Customers will need to protect themselves by taking control over their usage in order to make sure they are not overcharged.

    Share
    1. I agree about the usage !!
      Check out the new Android app: MyCalls , it will assist you monitoring your usage (Voice , Data , SMS) @ http://www.MoGoYo.com

      Share
  7. What about feature phones & plain old RTSP streaming? I’m looking for some statistics regarding bandwidth usage in RTSP video and would thank any leads.

    –Ran

    Share
  8. Have you heard about emerging technologies such as WiMax or 4G? Throughout your article you are assuming AT&T and other providers to stick with 3G. Once we have 4G, those 7GB/month figures look peanuts. And your graphs just show growth of HTTP streaming in %age – what is the base? Growing from 1MB to 100MB is 10000% growth rate – however if HTTP downloads/browsing are already 10GB then streaming doesn’t look that menacing – does it?

    Share
  9. Just got 20meg bb installed, now I can stream HD video off the net. This is the future people!

    Share
  10. “Unlimited” doesn’t make much sense unless the customer bit-rate is very low compared to the service provider’s capacity.

    Even consumer-grade home ISPs are having issues with “unlimited” traffic.

    A solution is to charge a reasonable “initial unit” charge that covers most customers, and a lower “incremental” unit charge for every unit after that. Whether that “unit” is just a byte-count e.g one terabyte, or whether it’s some combination of byte-count and time-of-day or quality-of-service (e.g. priority, latency, etc.) is a matter for the service provider and the market to decide.

    Share

Comments have been disabled for this post