Mobile data usage among different subscriber clusters
High-data users driving mobile traffic
- A significant proportion of traffic is generated by a limited number of users, while application mix changes across subscriber clusters.
- Video consumption was the dominant activity across every subscriber cluster, from light to extreme users.
- Social networking was the second most-consumed activity, with the highest share in the moderate to medium user groups.
Traffic measurements from mobile networks in two advanced mobile broadband markets show similarities and differences in application usage among different subscriber groups.
The analysis is restricted to data consumption on devices over cellular networks, and subscriber groups have been clustered based on their monthly data usage. It is based on data from traffic measurements in two commercial 4G and 5G networks in Europe and North America.
One-tenth of users generate 70 percent of traffic
The distribution of subscribers across different clusters and their data consumption varies from market to market, mostly depending on available data tariff plans. However, the traffic contribution of the top percentile of users (in terms of data consumption) is usually very similar. In both networks sampled, the top 10 percentile of users generated around 70 percent of the total traffic. In the North American network, users consuming over 20 GB per month represented only around 14 percent of all users, but generated 80 percent of the total traffic. A similar pattern was found in the European network, where users with consumption over 20 GB per month represented around 17 percent of all users, but generated 81 percent of the total traffic. Users with a monthly data consumption of over 50 GB represented only around 5 percent of users in the North American network and 7 percent in the European network.
Light consumers of data, those consuming less than 5 GB per month, make up 63 percent of all users in the European network. Among these users, a significant share of traffic, 16 percent, comes from communications services (messaging, VoIP, video calls and so on) and web browsing, while over 30 percent of traffic comes from a long tail of various apps.1
Application mix and traffic share in sampled networks
When analyzing the application mix and share of traffic in the sampled networks, it should be considered that these might not represent the absolute shares of the total traffic, as some traffic could not be classified. For example, the absolute share of video traffic is presumably higher across all subscriber clusters, as part of it is included in the category “Other”. However, it remains true that analyzing the relative changes in application mix over subscriber clusters provides insights into different data consumption patterns.
Video consumption: The dominant activity across all subscriber clusters. Intense and extreme users have the highest percentages of video consumption, accounting for over 60 percent2 of total traffic in the sampled networks. Share of video increases by more than 20 percentage points when comparing light to intense users in both networks.
Social networking: The second most-consumed application after video. The highest share of social traffic is in the moderate to medium user groups in both sampled networks.
Audio: There is a difference between the sampled networks, where the North American network share is 2–3 percent across subscriber clusters, while it is less than 1 percent in the European network.
Gaming3 and software downloads: These two categories represent a relatively low percentage of traffic, but with an increasing share of traffic with higher data consumption for both categories. For gaming, it is below 1 percent across subscriber clusters up to intense users, with an increasing percentage share for clusters with higher data consumption. This suggests that extreme users are more likely to engage in downloading software and gaming compared to other clusters. The share of gaming for extreme users in the European network was around 3 percent and in the North American network around 2 percent.
In both networks, the share of traffic for software downloads, file sharing and gaming was significantly higher for the extreme users compared to all other clusters.
Traffic share increase for video-on-demand among high-data users
In both sampled networks, social media-generated video is decreasing, while Video-on-Demand (VoD) streaming services are increasing their traffic share across subscriber clusters when going from light users to extreme users.
North America: Social media-generated video is experiencing a decline in its share of video traffic, from 88 to 49 percent, while the share of VoD streaming services is increasing from 4 to 23 percent.
In the North American network, YouTube has the highest share of video traffic across all user groups, with light and moderate users having the highest percentage. This is followed by Facebook and TikTok up to and including heavy users. For subscriber groups with more than 50 GB per month of data consumption, TikTok has a higher share than Facebook.
Europe: Social media-generated video is experiencing a decline in its share of video traffic, from 93 to 71 percent, while the share of VoD streaming services is increasing from 1 to 17 percent.
In the European network, Facebook has the highest share of video traffic across all user clusters, with light and moderate users having the highest percentage.
In both networks, the Netflix share of traffic goes from 1–5 percent among light to intense users to make up around 13 percent of the video traffic among extreme users (over 100 GB per month).
Facebook and YouTube have the highest video traffic share in both networks across all subscriber clusters, with a typical joint share of 50–60 percent of total traffic.
Across both networks, Facebook has a significant percentage of video traffic share for all subscriber clusters, but its share decreases significantly with increasing data consumption. The YouTube share of traffic displays a similar trend, while TikTok shows a trend toward an increasing share of traffic with increasing data consumption.
Impact of video
Video is having a significant influence on data consumption and traffic volumes in advanced mobile broadband markets. This trend is being driven by intense and extreme users, who have the highest percentages of video consumption. Video traffic share changes across subscriber clusters when going from light users to extreme users, with social media-generated video reducing its share in favor of a higher share of VoD streaming services.
1. For example: email, location services, photo sharing, weather, presence, health or fitness.
2. Unclassified video traffic is part of the category “Other.”
3. Includes both app-based and cloud gaming.
4. ”Other” includes uncategorized traffic and traffic from services that have too small a share to be significant compared to the categorized segments in this figure. A large share of ”Other” is presumably video traffic.
5. ”Unclassified” includes video traffic that was not possible to identify as a specific service or has too small a share to be significant compared to the specified services.