How do communication service providers (CSPs) deliver a great customer experience? Not that long ago, a CTO or CIO would have focused on the company鈥檚 network operations center (NOC), looking at metrics like performance and downtime. And, the thinking goes, if the company is providing high performance and minimal downtime, the customer experience must be great.

But the customer experience might not be great, even when the data suggests the network is performing well. For example, normal network performance metrics might indicate that the network as a whole is working fine, but if a streaming service customer is experiencing lag 鈥� perhaps because of a single streaming server 鈥� the customer experience is not great.

In other words, today鈥檚 communications CTO or CIO (and, in fact, the entire the senior team) should evolve from a network operations perspective to a service operations perspective 鈥� or, to use the lingo of the CSP, from network operations center (NOC) to service operations center (SOC). The 鈥渋nside-out鈥� view of the NOC may not be enough to satisfy today鈥檚 demanding customers and networks increasingly stretched to their limits. What鈥檚 needed in today鈥檚 world is an 鈥渙utside-in鈥� view to understand the customer鈥檚 actual experience with the network 鈥� moving to 鈥渘etwork enabled services and experiences.鈥�

This is both a mindset shift and an operational shift that harnesses the power of the enormous amounts of data that CSPs amass through digitalization and advanced algorithms to transform their networks into automated, self-healing, and self-optimizing systems to deliver intelligence that goes well beyond rule-based analysis and workflows. By analyzing myriad events, measures, and alerts, and deriving patterns from them, the SOC approach minimizes downtime and improves the customer experience. Put another way, today鈥檚 challenge is about shifting the focus from alerts about problems to opportunities to improve services.

CX: A Component of Operations

Evolving from NOC to SOC is better articulated as a transition to a unified service assurance solution that makes customer experience management an integral component of operations. A unified assurance solution continuously evaluates traditional elements like fault rates and network performance, but goes a step further to turn the network into an automated, self-healing system. 听

A unified assurance solution addresses all three layers that make up the customer experience through a single pane of glass:

• Network layer: This layer is what NOCs have traditionally monitored and protected, but in a service operations framework, these functions become more automated and autonomous:

1. Network outages caused by the failure of nodes or communications
2. Alarms and alerts that reveal exceptions and signal issues
3. Reactive processes that respond to a problem or situation
4. Key performance indicators (KPIs) that monitor and measure the performance of network elements and other devices 听

• Service layer: When companies make the switch to a SOC mindset, they consider new measures for monitoring the customer experience, including:

1. End-to-end service monitoring that evaluates service at the two ends of the communication and across chains of systems
2. Service inventory referring to the repository of internal services within communications
3. Key quality indicators (KQIs) including accessibility, retainability, mobility, and integrity

• Customer experience layer: This layer provides the holistic view of network and interaction experience touchpoints to improve customer satisfaction and Net Promoter Scores:

1. Network experience refers to the subscriber鈥檚 level of satisfaction with core services offered by the CSP; network experience is the result of the network service layer.
2. Interaction experience refers to the subscriber鈥檚 experience with other business systems and services (i.e., bill payment and customer service interactions).听

To deliver on the power of a Service Operations Center approach and provide this kind of unified assurance transparency, a digital solution must be built on these pillars:

• An automation-first approach so that workflows can be redesigned to embrace automation. For example, zero-touch provisioning enabled through an OSS and/or service/domain orchestration systems results in right-first-time orders and reduces cycle times significantly.
• Closed-loop automated workflows that allow the system to quickly detect, diagnose, and recover from issues that affect service quality or the customer experience without human intervention. This makes systems 鈥渟elf-healing鈥�; instead of alerts that warn CSPs of trouble and the need to intervene, they get notifications of problems averted or autonomously repaired.
• Open standards implementation to ensure interoperability across multi-vendor services.
• A data ops foundation with clear frameworks for the creation, lineage, and governance of data (the system is only as good as the data that populates it).
• Smart selection and analysis of network interface data provides an independent 鈥済olden鈥� version of the truth听鈥� but can be expensive. Identification of primary source(s) for the required measures combined with sampling strategy (for example, user plane data processing for VIP subscribers only) will help optimize infrastructure requirements.
• Artificial intelligence and machine learning to power data-driven, intelligent decision-making. For example, AI/ML applications leveraging the analytical data lake or fault management system(s) can provide automated alarm/event correlation and anomaly detection.听

The benefits of digitalization for CSPs based on these pillars are substantial:

• Lower fault rates and shorter mean time to recovery (MTTR)
• Zero-touch handling for significant issues (self-repairing networks)
• Improved service quality for customers
• Managed CX improves customer satisfaction (CSAT) scores and reduces churn
• Enables hyperscaling for next-generation digital operations using 5G, Internet of Things (IoT), network function virtualization, and cloud migrations

Next steps

91快活林 works with many clients in the telecommunications space to help them take advantage of the power of digitalization and artificial intelligence and make the evolutionary move from a NOC-centric perspective to a SOC-centric one. That experience has generated useful insights on the challenges involved:

• Standardization: Standardization of definitions and key measurements is a prerequisite. Network elements across radio, transport, core, and IP multimedia subsystem domains can be sourced from multiple vendors like Nokia, Ericsson, and Huawei. Without standardization, CSPs can suffer from service experience frictions and won鈥檛 be able to gain a unified perspective on all network elements.
• Since many CSPs have siloed systems and data for their network elements (radio, transmission, core), operations support systems (fault, performance, incidence, and others), and business operations (like CRM systems), moving to a SOC approach requires correlation of all these different silos of data.
• Availability of unified data allows addressing queries like impact of network alarms raised on, for example, a platinum subscriber鈥檚 jittery voice call. Further, unified data allows performance benchmarking and trending for peer network elements and subscribers. Once data is available throughout the enterprise 鈥� i.e., when silos have been removed 鈥� CSPs should calculate and remediate the impact of network data (i.e., network alarms, performance alarms) on related services.
• The process will create massive data lakes which need to be made navigable and monetized with infrastructure beyond dashboards and visualizations. For example, descriptive analytics capabilities summarize and visualize the alarms/alerts by individual network elements and their hierarchy. Going further, the advanced analytics approaches allow correlation, prioritization of alerts for immediate action(s).
• Finally, it takes a lot of human expertise (plus their associated time and costs) to understand and handle the brave new world of big data analytics. For example, a 15-minute slowdown in live streaming during a major global sporting event can trigger customer dissatisfaction and multiple complaints. The root cause of this problem may be buried within some combination of XDRs, logs, measures, alarms, and external servers鈥� latency performance. Data sourcing, processing, algorithmic correlation, retention all must be synchronized to support good, fast diagnostic analysis and to predict future occurrences.

But these challenges can be met with help from a knowledgeable partner. 91快活林鈥檚 experience in this realm has reveal several key success factors and levers for transformation:

• Maximum reuse of hardware data infrastructure to help manage the complexity and cost of this transition effectively.
• Minimize replication of data assets across NOC and SOC to reuse and reduce overall costs. For example, network probe data can run into tens of terabytes per day even for mid-tier operators.
• Continuous evolution of AI/ML applications to ensure ongoing relevance while data and AIOpS alignment leads to success. AI/ML solutions deliver the best ROI using a continuous integration and deployment approach that models performance validation, retraining with latest data, and monitoring for degradation.
• Robust data operations combined with governance will form strong foundation for the advanced AI/ML Operations-driven SOC. Access to data that is validated, of the right quality, in the right format, and at the required frequency and level of aggregation for specific use cases is key to success of the data driven SOC.
• Use APIs for northbound and southbound system integration to create opportunities for closed-loop workflows.
• Optimize infrastructure requirements through smart selection and sampling for sourcing high-volume data (such as user plane data for high-value customers).

The bottom line: Customer experience is becoming much broader 鈥� and more important 鈥� than network performance. But today鈥檚 technologies can give CSPs a powerful new way to harness massive amounts of data and use it in multiple ways to improve CX, differentiation, and growth.

听Learn more about 91快活林 and digital networks here.听

听