Advanced Television

Study: AI to limit operator revenue leakage

December 11, 2023

A study from Juniper Research, specialists in telecommunications markets, has found the average revenue leakage per 5G roaming connection will decrease from $1.72 (€1.60) to $1.20, as operators leverage AI-based segmentation. Revenue leakage refers to the value of services provided but not monetised.

This reduction in revenue leakage will be driven by the implementation of AI-based segmentation solutions to monetise data-centric users. Specifically, this approach allows operators to reduce 5G standalone revenue leakage through the improved allocation of resources and new pricing; reflecting its higher QoS (Quality of Service). The difference is that 5G standalone networks leverage the 5G core, whilst 5G non-standalone relies on 4G infrastructure.

AI Crucial to Developing Advanced Segmentation Solutions

AI-based segmentation will enable operators to better monetise emerging roaming services; using machine-learning models to detect and differentiate traffic types and segments in real-time.

Research author, Alex Webb, commented: “AI-based segmentation will differentiate enterprise traffic by use case; enabling premium billing of mission-critical 5G standalone connections, thus reducing revenue leakage.”

Non-standalone and Standalone Networks Must Be Monetised Differently

The report recommends operators implement AI segmentation tools to help reduce revenue leakage from 5G roaming on standalone networks. The higher throughput and lower latency offered by these networks needs to be reflected in operators’ pricing.

By separating standalone from non-standalone roaming traffic, operators will be able to apply individual pricing strategies for each of these networks; ensuring pricing reflects QoS. Operators must utilise these tools to identify enterprise traffic suitable for use case-dedicated network slices, as this reduces revenue leakage, by optimising network resource distribution.

Categories: 5G, AI, Articles, Mobile

Tags: , ,