To understand some challenges facing the communications industry and how our customers create value by analyzing large volumes of data, I spoke with Raquel Katigbak - an IBM business solutions executive specializing in the industry.
Mike: Raquel, this appears a good time for consumers in the communications industry.
Raquel: Yes, we are witnessing tremendous innovation, meaning, individually we have more choices than ever before. We live in an age of the empowered consumer where we can choose to connect and communicate from an array of handsets, smartphones and tablets. We have vendors and platforms offering countless mobile applications. And, we can choose to connect via many different carriers - both physical and virtual companies.
Mike: So how are CSPs (Communications Service Providers) coping with the challenges created by all that choice?
Raquel: Dow Jones’ Telecommunications Index tracks 74 companies with a combined market capitalization of 2.18 trillion dollars. In three of the previous five years the index contracted - overall it is down from 190 five years ago to 150 today. While the recent past has been challenging for many in the industry, the situation is not uniformly bleak. Focusing on just one factor - the rate at which customers depart for a competitor’s services (churn) - can generate profit. To illustrate, in the fourth quarter of 2011 Verizon reduced churn rate by ten percent to just 1.2% to lift average revenue per unit by 2.3%.
Mike: So what drives customers to churn?
Raquel: The most important factors determining customer loyalty are quality of service (QoS) and perceived value - see here for an analysis.
Mike: I understand that communications infrastructures are complex, but I expected these to be comprehensively equipped with systems’ management capabilities. Why is delivering high quality of service difficult?
Raquel: You’re right, communications infrastructures are complex - they are a system of systems. Their management systems, deployed within each sub-system, cannot see over their horizon in to neighboring sub-systems. Systems’ management throughout the entire network is fractured, not comprehensive. This problem is compounded by data retention issues. The management systems within each sub-system retain data only for the time needed to detect immediate problems - once that need is satisfied, they purge.
Mike: So an obvious starting point is integrating data from different management systems to build a comprehensive view and manage data for longer than retained in the underlying systems?
Raquel: Yes, that’s a good start, and by retaining data for months or years the CSP can identify trends or patterns of apparently minor problems which, over time, escalate to the detriment of service quality. For example, one North American CSP has five years’ of call detail records accumulated within their Netezza appliance.
But, this solves only half the problem. To create business value and become more competitive we augment network data with customer information. By integrating these two data sets a CSP understands relevant and important information about a customer: their location, their device, their behavior, how much revenue they bring to the CSP and the quality of service they experience - every time they access the network. Creating this network intelligence allows CSPs to answer such questions as “did we have appropriate capacity to respond to the customer’s need?” and “where will investment in new infrastructure create greatest profits?”.
Mike: Can you provide a real world example of a customer using IBM’s Netezza-based Network Analytics Appliance?
Raquel: I’ve recently worked with a large European CSP with a complex mobile network built on a high number of network components from multiple vendors. A jumble of systems and data marts could not provide an end-to-end view of their network. Their understanding of network performance and quality was 30 days old, leaving them no opportunity to affect and influence customer experience in a timeframe that would make a difference. Now with the Network Analytics Appliance they have near-real time visibility into network quality and performance. Their analytics relate these to customers, services, handsets and network locations and this understanding informs how the CSP manages individual customers, responds to their needs and drives a profitable business.
Mike: How long did it take to implement this solution?
Raquel: We ran a formal two week field trial. The appliance was receiving the CSP’s data within three days of being delivered; we then took requests from the customer and created new analytic queries to demonstrate the business agility created by this solution. Moving to production was straight-forward with no disruption to their business.
Mike: Where next for this approach?
Raquel: Similar QoS/network insight problems exist in fixed line communications - our experience and learning from mobile communications are relevant to these opportunities.
See here for more information.