9 Comments

Summary:

After rising eightfold through the next several years, the mobile data demand growth rate is set to slow says ABI Research. Wi-Fi offload will help, but there are two flaws with ABI’s forecast: Globally, feature phones still reign and our reliance on the cloud will grow.

gold tidal wave

The mobile data tsunami we’ve watched build over the past five years will crest in 2015 and the world will consume a whopping 107 exabytes through mobile networks in 2017. But the demand wave will grow slower in 2015, ABI research said on Wednesday, because “needless burden” on cellular networks will be addressed through Wi-Fi offload and smarter devices that can take advantage of free hotspots.

Although ABI predicts that monthly mobile data use is expected to jump eight-fold in the next five years, the rate of growth after such time will slow:

“It looks like 2015 will be the last year when the traffic volume will grow by more than 50% annually. And that will happen despite of the fact that the monthly average per wireless subscriber, worldwide, will increase to almost 1.5 gigabytes by the end of our forecasting period.”

Wireless offload is a topic we’ve addressed here many times in the past; there are definitely huge opportunities to reduce mobile data demand by supplementing mobile networks with Wi-Fi. The wireless industry is starting to embrace this solution, but it isn’t expected to account for a huge demand reduction for cellular data: Cisco estimates that only 22 percent of mobile traffic will flow through Wi-Fi in 2016, for example.

Like Cicso, I suspect ABI is over-estimating the effect of Wi-Fi offload, but that’s only half of the equation. There’s simply no reason to believe that the rate for mobile data demand will decrease starting in 2015. Here’s why:

First, we’re still a world of low- to no-data required feature phones. With roughly 7 billion people on the planet, only 16 percent at the end of last year had a smartphone, estimates Tomi Ahonen. Far fewer have tablets. But in the next three years, the overall user base for data is sure to grow as smartphones  decline in price and network infrastructure improves worldwide. That means more demand for some time to come in my opinion.

Second is the problem of “not knowing what we don’t know.” By that, I mean we simply can’t predict what mobile device activities will require data, nor how much of it. We’re moving from email to social networking apps and local video to sharing or downloading digital media over wireless networks.

Simply put: Our reliance on the mobile Internet is only just beginning and the web-based activities three years from now can’t be determined. Again, Cisco seems to agree, suggesting that cloud services and apps will account for nearly three-quarters of mobile broadband use by 2016.

Without a doubt, Wi-Fi offload will help matters. But we have so few solutions trying to combat a near-infinite number of mobile use cases — many we haven’t yet envisioned — for us to assume that mobile demand will slow three years from now.

  1. A lot depends on how mobile data is priced and the options that are available to control that by smartphone users. Having a smartphone use wifi where available for data can save a lot where there is wifi at home and at work and at many public places. It’s faster than mobile data will ever be and doesn’t cost extra.

  2. Ram Krishnan Friday, July 27, 2012

    Well,

    Mobile traffic is definitely slowing – according to Vodafone’s Interim Report. Vodafone is a reliable bellwether and traffic growth in its advanced geography, viz., Europe is definitely slowing down.

    See
    http://www.vodafone.com/content/dam/vodafone/investors/financial_results_feeds/ims_quarter_30june2012/p_ims_30june2012.pdf

    Network traffic grew from 49PB in Q1 11/12 to 62PB in Q1 12/13 – year-over-year growth of only 26.5%. VF indicated in October of last year that its traffic management techniques are working quite nicely and traffic was no longer growing exponentially. I completely expect this dynamic to continue

    1. Thanks for the data, Ram. How much of Vodafone’s traffic information apply to emerging markets and other areas behind the mobile broadband curve though?

  3. I hear the arguments regarding continued explosive growth in cellular wireless demand, but I am intrigued by a recent article in the Economist which describes how the conversion to digital television suddenly makes the “white space” between VHF channels available for WiFi at much higher broadcast power. Should this make its way through the regulatory process, we would have true long range metro WiFi, and wireless demand would taper off.

    1. Yup, we’ve covered white space networks since 2010, but not much has happened there yet. Here’s some info: http://gigaom.com/2010/09/22/all-you-need-to-know-about-white-spaces-broadband/

      I think the issue there will be that “white space” spectrum will be unlicensed, if I recall correctly. That means it’s similar to Wi-Fi in that anyone can essentially use it. I’m simplifying the explanation, but the point is that it could be used in multiple ways by non-operators.

      1. Kevin, your response that Gigaom is “following” white space sounds like “whistling in the graveyard” to me. You acknowledges that essentially anyone could potentially appropriate the space, including high power metro Wifi operators. The prospect of white space as the most logical and efficient wide are use of white space, as opposed to a “free for all,” is mobile’s worst nightmare… As an Intel alumni veteran of Intel SIG standards wars, I think regulators are likely to see the merit of rationalizing use of white space with higher power metro WiFi, and the Intel Labs folks should jump on it. Let’s hope that my friends there cover that base.

        1. David, I don’t suggest, nor advocate a “free for all” with white spaces. I strongly hope standards for this unlicensed spectrum are followed so that white spaces are used widely in a fashion similar to other unlicensed spectrum use; namely that of Wi-Fi and Bluetooth.

  4. I would expect aggregate cellular data traffic growth to fall below 50% in 2014, not 2015.

    Much of the developed world is now well below 100% growth (for example most European operators), and even operators like China Mobile are reporting just 77% growth, while WiFi accelerates much faster.

    In any case, neither the cost structure of developing markets, nor the mathematics of traffic statistics, is likely to support the sort of ultra-high usage growth seen in recent years. Users will have (comparatively) cheaper/weaker devices, on thinner 3G/4G networks, and more restrictive data plans – mostly prepay.

    $5 ARPU doesn’t buy much spectrum/subsidy/network capacity.

    I wrote a long blog post on the traffic growth myth recently – http://disruptivewireless.blogspot.co.uk/2012/08/mobile-data-traffic-growth-thought.html

    Best regards

    Dean Bubley
    Disruptive Analysis

Comments have been disabled for this post