Liberty Latin America has launched AI-based Cx Tech, a program that aims to understand how technical experience influences customer experience and churn. Traditionally, network engineering addressed user experience issues by finding and troubleshooting potential broadband disruptions using metrics like Codeword Error Rate. The AI-based Cx Tech program enhances this by using advanced analytics, big data, and machine learning to find critical events and metrics in HFC and FTTH networks.
However, telcos often struggle to correlate direct network KPIs with churn because the KPIs are just the beginning of the problem. A customer's decision to churn is also heavily influenced by how effectively the issue is addressed afterward. This requires a shift towards a more comprehensive approach in network intervention strategies.
Different customers have different usage profiles, and the same kind of network issues can affect them differently. Therefore, customer-centric prioritization is essential to address these variances in impact effectively. The AI-based Cx Tech program incorporates this perspective, ensuring that interventions are tailored to the specific needs and experiences of diverse customer segments.
A key tool from this program, Lighthouse, is designed to optimize network interventions to maximize the impact on reducing customer technical calls. Recently, Lighthouse's HFC network segmentation was enhanced with generative AI call classifiers from customer transcripts and threshold optimizers from machine learning models. By using Bayesian Optimization with Gaussian Processes and Genetic Algorithms, this improvement has aligned segmentation more closely with customer experience, thereby improving service quality and satisfaction.
Besides improving the understanding of network performance, there is a need for a proactive network intervention redesign. Such a redesign enables prompt reactions to customer pain points, further enhancing the effectiveness of technical solutions and ultimately reducing churn.