Sandvine has just released its latest half-yearly Global Internet Phenomena Report. The reports are based on data collected by Sandvine from among its 250-plus customers spanning, well, the globe. Several items jump out this time, but two are especially interesting.
First, video continues to crush the numbers. Not that this is any surprise, since the dominance of video has been in the forecasts for years. What is surprising is the odd bedfellows that now account for 50+% of downstream traffic on wireline broadband in North America: Netflix and YouTube.
Netflix just won’t let up – in traffic share, share of mind or share price. Despite the ham-fisted attempts by ISPs like Rogers and Bell to make OVDs expensive for its customers by screwing them with data caps, Netflix’s traffic share is still holding at about the same level as six months ago – currently 31.6%. I’ve written a lot about Netflix over the last two or three years, much of that in defence of our right to choose what we watch in Canada.
I’ve also noted Netflix is an important bellwether for the future of online video – and thus for the future of what we call TV. What sets it apart from almost everything on cable is that it’s an interactive service in which customer feedback and loyalty to the brand play a huge role. And that makes comparisons based on sheer revenue misleading. For example, my colleague Dwayne Winseck has pointed out that Netflix accounts for a mere 1.8% of TV revenues in Canada – implying Netflix will do little to change how video is packaged and consumed over the next few years.
I disagree. Once TV has really moved to IP, Netflix will have tremendous first mover advantages and its share of time spent online will force the old media guys to play catch-up. All that will, of course, take years, not to mention a serious effort on the part of the CRTC to stamp out the undue preference our incumbent ISPs continue to give their own content. (Sandvine says Netflix has also become a huge bandwidth consumer in the UK only two years after its launch there.)
The other winner in the current video sweepstakes is YouTube, which now accounts for 18.6% of North American downstream traffic – adding up to over 50% of the total with Netflix. There’s another prognostication the TV dinosaurs got wrong. Ever since Google bought YouTube in 2006, I’ve heard people from legacy media poo-pooing YouTube as a haven for dancing cats with zero production values and no future up against “professional” video content. For all that Google has been trying to make its platform more professional, its appeal still lies in its user-generated funkiness, not the greater availability of content in 720p.
These shares for Netflix and YouTube are all the more remarkable given how many kinds of video have become available on the public Internet. As Cisco points out in its 2013 VNI forecasts for IP traffic, almost every single category of video is climbing sharply. Here are a couple of examples:
- Internet video to TV doubled in 2012. Internet video to TV will continue to grow at a rapid pace, increasing fivefold by 2017. Internet video to TV traffic will be 14 percent of consumer Internet video traffic in 2017, up from 9 percent in 2012.
- Video-on-demand traffic will nearly triple by 2017. The amount of VoD traffic in 2017 will be equivalent to 6 billion DVDs per month.
Think you’ve got video overload? You ain’t seen nuthin yet…
- By 2017, it would take an individual over 5 million years to watch the amount of video that will cross global IP networks each month.
For the last couple of years, Cisco has also been forecasting a steady decline in the relative quantity of peer-to-peer filesharing traffic sent over the Internet. Sandvine confirms this trend: P2P now accounts for less than 10% of total daily traffic in North America, whereas five years ago it accounted for over 31%.
Cisco says P2P will increase in global volume from about 6.2 exabytes a month in 2012 to about 8.7 exabytes in 2017, which works out to a CAGR of 7%. (Cisco defines P2P filesharing as traffic from all recognized P2P systems such as BitTorrent and eDonkey, as well as traffic from Web-based filesharing systems.)
Piracy is not on the increase, however, because other traffic is increasing at a much faster rate than filesharing. Other forms of Internet video traffic are increasing by a CAGR of 29% – over four times the growth rate of filesharing. The reason for this drop has a lot to do with the advent of legit, on-demand streaming services like Netflix. And it so happens that trend fits nicely into Sandvine’s business model, centered on what they call network policy control technologies (including what the rest of us sometimes call deep packet inspection or DPI). Here’s the view from Dave Caputo, Sandvine’s CEO:
“For the first time ever, peer-to-peer filesharing has fallen below 10% of total traffic in North America, which is a stark difference from the 60% share it consumed 11 years ago. Since 2009 on-demand entertainment has consumed more bandwidth than ‘experience later’ applications like peer-to-peer filesharing and we had projected it would inevitably dip below 10% of total traffic by 2015. It’s happened much faster. This phenomen[on], combined with the related rise in video applications like Netflix and YouTube, underscores a big reason why Sandvine’s business has grown beyond traffic management to new service creation.”
I was discussing these trends over coffee yesterday with Mike Zajko of the University of Alberta, who’s currently doing research on several aspects of IP traffic. As Mike pointed out, our incumbents are now going to need new demons – other than pirates – to blame for the alleged traffic congestion on their networks.