Boosting NPS beyond network KPIs with agentic AI

Moving from network metrics to real drivers of customer satisfaction

CSPs have long relied on network KPIs and surveys to understand customer experience. But these approaches often capture only part of the picture. Customer satisfaction is shaped by many factors beyond network performance, making it difficult for operators to identify what truly drives Net Promoter Score (NPS) or act on it with confidence.

The Catalyst, Boosting NPS beyond network KPIs with agentic AI, explores how agentic AI can transform how CSPs measure and improve customer experience. Rather than relying on fragmented, periodic data, the project introduces a continuous, data-driven model that links operational performance directly to perceived customer outcomes.

The challenge of measuring real customer experience

Traditional NPS measurements depend heavily on surveys and limited datasets. These methods can miss the wider context behind customer sentiment, leaving CSPs to make retention and investment decisions based on incomplete or delayed insight.

The underlying issue is fragmentation. Customer experience is shaped by network performance, service interactions, usage patterns, and external factors. Without a unified view, it is difficult to understand how these elements interact or which ones matter most to different customer segments.

As Jayraj Roodurmun, Principal Officer at Mauritius Telecom Ltd, explains, the Catalyst takes a “disruptive and game-changing approach” by grounding NPS evaluation in a wider set of real-world variables. He notes that agentic AI can capture a broader range of parameters and enable intent-based management in highly dynamic operating environments.

A new approach to NPS optimization

The Catalyst introduces a multi-domain NPS estimation and optimization framework that combines network performance data, including CEI and QoS/QoE indicators, with customer care interactions, incident and fault management data, service onboarding and lifecycle data, device and service usage signals, crowdsourcing insight and offer positioning data. Together, these inputs create a more complete, real-time view of customer experience than survey-led approaches can provide.

AI and machine learning models establish correlations between customer experience indicators and NPS, while driver analysis identifies the factors that most affect satisfaction by segment.This gives CSPs a clearer view of which operational, care and service issues are most likely to influence loyalty and churn.

Agentic AI then supports continuous monitoring, reasoning and autonomous decision-making. Insights can be translated into automated or assisted actions across service onboarding, customer care, and network and service lifecycle management, creating a closed-loop approach to customer experience optimization.

From insight to action

The shift from analysis to action is central to the value of this approach. Instead of relying on manual intervention, the system supports continuous optimization across the service lifecycle, helping CSPs identify experience drivers for different customer segments and address them faster.

The project aligns with several TM Forum assets: IG1394 Telco NPS Management Framework v1.1.0 for NPS collection, correlation and optimization across journey, product, service and network domains; the Autonomous Networks Framework, from AN Levels L0 to L5, for closed-loop automation and cross-domain orchestration; Open Digital Architecture (ODA) for modular, interoperable components; and TM Forum Open APIs where applicable for integration across network, service and customer domains.

Driving measurable outcomes

The expected impact is improved NPS accuracy, measurable uplift in targeted NPS use cases and reduced churn through proactive experience management. For CSPs, the solution helps identify the true drivers of satisfaction across network, care and operational domains, so teams can focus on the interventions most likely to improve customer perception.

For the wider industry, the project contributes to cross-domain KPI correlation models that bring together CEI, NPS and operational data, while advancing the use of agentic AI in telecom operations and customer experience-driven autonomous operations. For end users, the result should be more consistent, reliable services shaped by real interactions and experience.

Why it matters

This Catalyst shows how CSPs can make NPS more predictive, operational, and actionable. By using agentic AI to connect data, insight and action, it enables a shift toward continuous, outcome-based customer experience optimization.

As networks and services become more complex, the ability to understand and influence customer perception will become increasingly important. This project demonstrates a practical path to stronger loyalty, lower churn, and differentiation through experience.

DTW Ignite 2026

Boosting NPS beyond network KPIs with agentic AI was showcased at DTW Ignite 2026 as part of the Catalyst program, demonstrating how agentic AI can help transform customer experience from reactive measurement to continuous, outcome-driven optimization.