ML-Based Data Quota Assignment for Wireless Subscriber Retention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cellular networks face challenges in managing customer relationships and expanding wireless networks effectively, particularly in identifying high-value customers and optimizing data usage quotas to prevent churn.
Innovation Solution
The use of machine learning-based techniques to collect and process data on user behavior, allowing network providers to estimate customer lifetime value (CLV) and predict churn rates. This enables targeted allocation of additional data usage quotas to high-value customers and strategic placement of wireless retail locations and network infrastructure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If bandwidth throttling is applied to users who exceed data quota, then network costs are reduced in the short-term, but user experience deteriorates and churn rate increases
Solution Approach 1:
The system performs preliminary actions by proactively identifying users likely to exceed their data quota before they actually do so. Machine learning models predict future data usage patterns and alert users in advance, allowing them to take corrective actions (such as purchasing additional data or adjusting usage habits) before throttling would occur, thus preventing churn while managing network costs
Solution Approach 2:
The system implements continuous feedback loops where user data usage patterns are monitored in real-time, predictions are updated based on actual usage versus predicted usage, and users receive ongoing notifications about their consumption patterns. This feedback mechanism allows dynamic adjustment of warnings and interventions, improving user experience while maintaining cost control
2Reliability
If additional data usage quota is allocated to high-value customers, then churn rate is reduced and user experience is improved, but network costs increase
Solution Approach 1:
The system applies local quality by differentiating treatment based on individual user characteristics. Machine learning models segment customers into different risk profiles and value categories, applying targeted interventions only to high-value customers who are likely to churn. This selective approach ensures that additional data allocations are concentrated on users where they will have the greatest impact on retention, rather than being applied uniformly across all users
Solution Approach 2:
The system dynamically changes parameters such as data quota allocations, warning thresholds, and intervention strategies based on predicted customer lifetime value and churn probability. High-value customers receive more generous data allocations and earlier warnings, while lower-value customers receive standard treatment, optimizing the balance between retention effectiveness and network cost
3Productivity
If machine learning models are used to predict customer lifetime value and churn, then resource allocation is optimized, but system complexity increases
Solution Approach 1:
The system segments the customer base into distinct groups based on predicted lifetime value and churn probability using machine learning models. This segmentation enables differentiated resource allocation strategies for different customer segments, improving overall resource allocation efficiency. The models process user data to create segments that can be targeted with appropriate interventions, balancing the complexity of modeling with the benefits of precise resource allocation
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for machine learning-based assignment of data usage quota to wireless customers. A method includes identifying a set of one or more features corresponding to a wireless subscriber. The method also includes determining, based on a first portion of the set of one or more features and using one or more first machine learning models, that the wireless subscriber is likely to exceed a quota of data usage allotted to the wireless subscriber. The method also includes determining a score that is indicative of an expected profitability associated with the wireless subscriber, determining that the score satisfies a threshold condition, and responsive to determining that the score satisfies the threshold condition, allocating an additional data usage quota to the wireless subscriber.