Summary
Viewing video content over the Internet is becoming more and more popular in Germany, as in many other countries. Thus, video content already accounts for the vast majority of data transported over the Internet. It is expected that the proportion of video content will increase significantly in the future. Long-term network capacity planning rests on these forecasts. But how reliable are the forecasts conventionally used and which demand-related developments influence the increase in data volume in particular?
To answer these questions, this paper examines the development of demand for video streaming services in Germany based on representative surveys conducted in Germany in 2015, 2016 and 2017 as well as the available range of streaming services and their business models. Three possible scenarios for future video consumption are derived from this. Within that we consider the evolution of codecs and adaptive streaming technology to arrive at a realistic estimate of the future data volume.
The results underscore the marked increase in the usage intensity of streaming services in Germany. Notably, the increase in usage can be attributed mainly to the increase in usage intensity of those, who already use online streaming services. Overall, the proportion of consumers in Germany relying solely on streaming services almost doubled from around 12% to around 21%. Meanwhile, the proportion of traditionalists, who do not use streaming services at all, remains relatively stable over the three years observed (33% in 2015 and 29% in 2017).The baseline scenario assumes that market saturation has already been reached today. It extrapolates the current intensity of use. The growth scenario continues the trend of the years 2015 to 2017 assuming all video content will eventually be viewed via the Internet. An intermediate scenario was defined between these two extremes. These scenarios were combined with the probable development of the required resolutions and codecs. It turns out that the usage intensity plays a significantly less important role for the amount of data to be transported than the required resolution. As long as the resolutions follow their current trend, our scenarios confirm the published expectations regarding the development of data volumes in Germany. If there were significant jumps in screen resolutions or if virtual reality formats were adopted faster than expected the forecasted data volume would be exceeded many times over.
Discussion Paper is available for download.
