It is a truism in the high speed access business that today’s “heavy users” are tomorrow’s “typical users.” And nothing will provide more confirmation of that theory than future consumer behavior, if and when the same content now available from linear video services shifts to Internet delivery.
Already, real-time entertainment is responsible for over 63 percent of downstream fixed network bytes during peak usage periods.
During peak period, real-time entertainment traffic also accounts for over 40 percent of the downstream bytes on North American mobile networks as well.
In a future market where significant amounts of former linear video viewing are shifting to over the top Internet delivery, the amount of bandwidth consumed by even “typical” consumers will increase by orders of magnitude.
Consumers who do not buy linear video services consume roughly an order of magnitude more data than consumers who rely on linear video services for most of their video consumption, the data suggests.
“Cord cutters” are U.S. ISP customers in the top 15th percentile of streaming audio and video usage. And while it is impossible to ascertain directly, Sandvine assumes those consumers are using streaming as a primary form of entertainment.
That suggests a high likelihood those consumers either do not buy traditional linear video entertainment services, or have access to such services but do not find them as compelling as over the top alternatives such as Netflix, Amazon Prime or Hulu.
“Typical subscribers” in the 15th to 85th percentile of data consumption likely are streaming on a regular basis, Sandvine says, and arguably represent “most” U.S. consumers in terms of behavior.
In other words, these days most consumers stream some amount of video or audio.
“Non-Streamers” represent the bottom 15th percentile of users who stream less than 100 MB of audio or video each month.
Those statistics have direct bearing on future market trends, when it will be possible for consumers to stream much of the content they otherwise would watch as part of a traditional linear video subscription.
The implication is that demand for Internet access bandwidth will skyrocket, even as buying of linear video services declines.
Assuming 1.5GB of data for each hour viewed, and assuming relatively equal high definition and standard definition content viewing, subscribers with “cord cutting behavior” could be consuming on average roughly 100 hours of video each month, or 25 hours a week or roughly 3.6 hours a day.
The top 15th percentile of video users actually consume the majority of monthly network traffic, while the bottom 15th percentile of users consume only 0.5 percent of data.
Again, that suggests the scale of future demand for Internet access data demand. If and when today’s linear video content is widely available over the top, the typical consumer will have behavior very similar to current behaviors of the heaviest users, at a minimum.
For the first half of 2014, mean fixed network (“average”) usage was 51.4 GB, a slight sequential increase from the 44.5 GB seen in in the second half 2013 report.
Over the same period, median fixed network (half of respondents use more, half use less) monthly usage also saw a small sequential increase moving from 17.6 GB to 19.4 GB.
But many ISPs predict 30 percent to 40 percent annual growth rates for aggregate demand, in 2014.
Mean Monthly Usage
Mean Real-Time Entertainment Usage
Average Hours of Streaming
Share of Total Traffic