Skip to main content

AVA - AI for video services

Use AI to uncover hidden patterns to enhance experience

Why is telco customer experience the top priority for CSPs?

Customer experience has been and is a battlefield for telecom market share. Mobile video consumption rises 100% every year (Insivia) and will account for nearly 80% of mobile data traffic by 2022. The cloud gaming market is reaching $8 billion in value by 2025 and the largest driver of growth across the video game industry is from mobile gamers. End users have less and less tolerance for QoE degradation - jitter/buffering/freezes etc.

If mobile operators are to convert the promise of 5G into new revenues from new services, they will need to deliver the consistently high network experiences which consumers have come to expect. But providing quality experiences will not be straightforward.

Approaches to network quality and network experience need to change because of the changes in market focus in 5G.

How to enhance telco customer experience for end-users?

To help the CSP address the challenge, Nokia developed its AI for video-based services solution which is part of Nokia's AI and analytics portfolio. This analyzes video/gaming performance data including encrypted video, to predict how Quality of Experience (QoE) will be affected by interference, congestion and coverage.

Which KPIs for video services telcos need to track?

Modeling techniques then link video/gaming QoE to critical business Key Performance Indicators (KPIs), including churn, Net Promoter Score (NPS) and revenue. The CSP is provided with recommendations on which parameters to alter for resolving issues quickly. Linking QoE to business performance also allows optimization efforts to be directed at the most valuable subscribers and locations.

Correlation of User Plane traffic and Layer 3 signaling enables prediction of 'freezes' and QoE. The solution is multivendor capable and can make use of a variety of data sources. Nokia offers flexible commercial models. 

Case study: how telco AI helped to fix issues with video for a tier 1 US operator

For a US CSP, the solution made 54 automated recommendations, allowing it to optimize the performance of the network by changing parameters. The worst performing cells were highlighted, as well as issues affecting the transport, core and content delivery networks.

How CSP benefited from using telco AI for video based services

The CSP gained significant benefits from the service. 

  • The main ones were a 59% reduction in NetFlix buffering. 
  • YouTube services that suffered playback delays of more than four seconds, were reduced by 15% - after four seconds of delay, most people abandon a video, so cutting this delay is vital to keep subscribers using the service. 

The CSP gained a better understanding of what drives video QoE and user behavior, as well as the ability for rapid troubleshooting and prioritization of issues. It can now also predict any bottlenecks across the network, whether they occur in the radio network, transport, the core or content delivery network (CDN). Because it improves the user's experience, AVS can have a positive effect on the CSP's Net Promoter Score.

Case study
US case study video analytics

Predictive Video Analytics

Ready to talk?

Please complete the form below.