Last month, Dell Technologies was the recipient of the FutureNet award for Most Innovative Application of AI to enhance Customer Experience, in recognition of a solution deployed by Dell Technologies and ISV partner Cardinality at O2 (Telefonica U.K.).
Network Customer Experience Solution
The solution, which O2 calls Network Customer Experience (NCX), uses machine learning to evaluate patterns in massive volumes of network data and then assign an “NCX score” to every subscriber on O2’s network. The NCX score, in turn, is used by multiple departments within O2 to inform decisions around various business priorities including customer churn reduction, network quality of service improvements, device upgrades and service upsell.
The most exciting aspect of the NCX solution is how it helps O2 achieve the “Holy Grail” of telecom data analytics, which is to correlate network events directly with the experience of individual subscribers. With the ability to stream over 15B real-time events per day, Dell Technologies SP Analytics with Cardinality feeds over 300 network metrics (also known as network KPIs) into its machine learning algorithm. The algorithm, in turn, evaluates the KPIs based on their relative impact on customer satisfaction to generate an individualized NCX score for each active subscriber. O2 spent many months surveying its customers to weight the relative importance of network KPIs and then used the survey results to train the algorithm. The result is that O2 can take the “temperature” of subscribers at any given point in time and take individualized actions to maintain loyalty and improve the customer experience. Among the groups within O2 that utilize NCX insights in day-to-day operations are Network Operations, Marketing, and the CTO Office.
The NCX solution offers benefits to telco carriers of all sizes and these benefits can be realized in both revenue gains and cost reductions. Some examples¹:
- A net present value ROI for the overall project of $2M over 2 years for an initial project investment of $4.5M.
- Extending that same ROI over 6 years, cashflow generated from the same project could exceed $40M, a significant payback in cost savings and revenue from the initial $4.5M investment.
- 15% decrease in the cost of retaining subscribers. The total cost savings will vary from carrier to carrier based on number of subscribers, cost of retention and geographical market saturation.
- 20% increase in upsell, increasing total revenues to offset the decline in average revenue per user (ARPU) seen across the industry.
- Customer churn can typically be reduced by 1% to 2% over a 6-year period, which can be significant for large carriers with hundreds of thousands of subscribers.
- Complementing churn reduction is the number of subscribers that can be added as the result of implementing a project like NCX. Operators with 150,000 or more subscribers can see net new subscriber adds of over 100,000 over that same 6-year period, directly attributable to the insights from data analytics and subsequent actions taken to build the subscriber base.
The Dell Technologies solution at O2 is built on the following technology foundations:
- Dell Technologies’ approach to telco data analytics, which is a departure from traditional methods of delivering data-driven insights to customers. Rather than attempting to “own” the entire analytics stack—from data ingestion to storage, analytics, visualization, machine learning and automation—and thereby locking customers into monolithic vertical solutions, Dell Technologies gives customers choice. Do you prefer Tableau or Kibana for visualization and reporting? Zeppelin or Jupyter for computational analytics? Software-defined storage or a standalone storage system for your data lake? With Dell Technologies, customers can pick and choose the tools they are familiar with and avoid having to retrain staff and integrate unfamiliar technology into their analytics environments. By validating Dell Technologies hardware, storage and networking infrastructure with a broad ecosystem of partners and toolsets, we make it possible for customers to deploy analytics solutions in an open and disaggregated way while ensuring they get the performance and data management required to create innovative use cases, all while maintaining control and ownership of their data.
- The breadth of the Dell Technologies product and solution portfolio for analytics, which features a range of products that are purpose-built for a variety of analytics, machine learning and deep learning applications (such as the PowerEdge C4140, equipped with NVIDIA Tesla GPUs to unlock deep learning capabilities for applications ranging from computer vision, natural language processing and autonomous driving), and delivered as part of fully integrated and optimized Dell Technologies solutions that combine hardware, software and services to eliminate guesswork and reduce risk for customers, allowing them instead to focus on deriving business value from their data.
- Cardinality’s telco-optimized Perception platform, engineered to stream data from network probes and other telco data sources and then perform real-time data prep, cleansing and analytics on that data, storing the results and making them available to business users in the form of reports and visualizations, and to data scientists for further analytics and training of machine learning algorithms. Perception is Kubernetes-based and can run on-prem, in the cloud, and in hybrid environments.
- Dell Technologies ability to deliver and support end-to-end solutions for customers globally.
- O2’s ability to innovate on the joint solution and provide cross-functional value to business stakeholders looking to make data-driven decisions.
With award-winning solutions like NCX and a growing ecosystem of best-of-breed partners like Cardinality, Dell Technologies is excited to become the preferred IT infrastructure provider for telecom as the industry evolves toward 5G and data-centric business models built on disaggregated, cloud-based architectures.
¹ “O2 UK: Driving Customer-Centricity in Network Analytics”, STL Partners, November 2019