We have all seen the infographics: a proliferation of connected devices is fueling explosive mobile data traffic growth, with mobile accounting for ever-higher proportions of the total volume of data traffic.
But is this really true? This article challenges two pieces of received opinion in data traffic forecasting: that mobile’s share of traffic will continue to grow, and that it is the ever-expanding range of smaller connected devices that is driving internet growth.
Assumption: Mobile to outgrow fixed data traffic
It is not too early to suggest that mobile share of data traffic will actually peak, at least in developed economies.
This statement is clearly not true for all operators. To take an extreme example, Telefonica Brasil’s (Vivo’s) fixed broadband data traffic was growing at an organic rate of about 90% per annum at 1Q 2016, whereas for mobile, the rate was 22%.
Individual operators have differing priorities, but this premise is also not true for all countries, particularly the most advanced. Japan has among the highest mobile usage levels per capita of any country.
However, in early 2015, fixed data volume growth, as recorded by the Japanese Ministry of Internal Affairs and Communication, overtook that of mobile, indicating that mobile has a declining share of internet traffic.
Part of the reason is that Japan has historically had low overall (fixed plus mobile) internet usage per capita compared with peer-group economies. However, Hong Kong also has high mobile usage and “quite unlike Japan” extremely high average fixed broadband usage (160GB per month, on average), and according to regulator OFCA, fixed data growth rates there still overtook those of mobile in 2016.
Our view is that the rates of growth in advanced economies will converge and that, for the end of this decade, fixed will show marginally stronger growth in most markets.
Even in middle-income emerging markets, the two rates will close to almost parity by the end of the decade. Only in the least developed markets will mobile continue to grow in share.
We believe this will happen because normal growth in average usage on an already-installed base of devices is about 30% per annum. Higher than normal patterns of growth are usually achieved through the introduction of new classes of devices.