SAS on telco data modeling: Avoiding 'contagious churn,' enhancing 'viral marketing'

SAS recently acquired marketing tools vendor AssetLink, which helps marketers in a variety of markets develop more intelligent marketing campaigns. In telecom, the company sees a particular interest in finding new ways to sell existing customers new things, as well as intelligently deciding where to build out networks.

Specifically, operators are seeking solutions that optimize the value hidden in both historical and real-time data coming through networks and OSS/BSS.

“In North America, we increasingly work with service providers that are keen to examine not only segmentation, churn and customer lifetime value but new things like social networking impact on their brands, or the relationships between customers so they can recognize group leaders and their influence on others in terms of buying products or switching to competitors,” said SAS’ Ken King, marketing manager for communications industry.

While data mining and analytical models have been around a long time, they are now being used in marketing campaigns to more proactively predict behaviors and adjust offers in real time reach out to customers who are the greatest influencers—or so-called “group leaders.”

“By analyzing the voluminous amounts of data now sitting in telco repositories, you can look at call patterns and build knowledge about who the leaders and followers are in certain groups, and what they do on their smart phones, PCs, set-top boxes and other devices,” said King. “ou can then do content categorization and start recognizing certain phrases and triggers that help you build taxonomy around what you learn—all with the intent to better understand not only who a person is and what his standing is in a circle of inter-connected people, but whether he is happy or not, and to what degree he is happy or not. Then you can move onto things like ‘sentiment analysis’ to predict triggers and indicators of what their actions are going to be, which helps an operator proactively respond.”

In short, operators want to do more to recognize marketing opportunities and to contain costs.

“Predictive analytics for looking at services and forecasting outcomes, as well as optimization of networks, are all growing in importance for operators that want to understand customers, services, networks and the problems related to each. They need early indicators of where troubles lie—whether in the marketing, the fulfillment of services, the fleet management or any areas where they can dedicate resources for improving the customer experience,” said King, noting operators no longer want to be responsible for “gluing together” different solutions from analytics players and OSS/BSS partners. “They want pre-integrated solutions, so we are focusing on partnering more and more with OSS/BSS players so that pre-integrated predictive analytics leverages the data they already have stored in databases and running through management systems,” added King.

For this reason, he expects that more partnering and integration will happen to help operators better analyze the data they increasingly collect for marketing, operations, regulatory and other business purposes. “Since they are collecting and storing data for many different reasons, they realize they should derive as much value as possible for the sake of profitability, operational efficiencies and costs savings, and improving customer experience,” added King.

As a result, in-database scoring and analytics are going to become more and more important to operators, which no longer want to pay for separate analytical data marts for different models they have to run for different purposes. And, they will increasingly seek analytics vendors that form partnerships with OSS/BSS providers and systems integrators to ensure initiatives like those for subscriber data management or policy management are leveraging advancements in modeling and in-database scoring without major investments on their part.