Viewpoints
Industry Insights
Circle of Friends
Social network analysis helps telcos better understand customers.
by Drs. Judy Bayer and Edouard Servan-Schreiber
Telecommunication leaders invest considerable effort to find reliable ways to analyze their customers'; behaviors and preferences. Having this type of data can allow companies to identify not only the types of products and services that interest their customers but also how they are used.
A behavior analysis can help validate how individuals interact with people inside and outside their social network, as well as how much purchasing power and influence they hold in their community. With this information at hand, telcos can more effectively cross-sell and up-sell products and services to subscribers and, even more importantly, keep them from churning to the competition.
However, the methods most used for analyzing and predicting behavioral trends in this high-tech, fast-changing industry often fail to fully harness the capabilities of modern-day tools and technology.
Social network analysis (SNA) is fast emerging as an important discipline for predicting and influencing consumer behavior. This methodology focuses on how relationships are built and how they contribute to influencing individuals within social groups. When coupled with the advanced information and analytic power of an enterprise data warehouse (EDW), SNA offers previously unimagined resources to determine how to successfully retain and grow profitable customers.
FAR-REACHING RESULTS
In the telco industry, key business problems are often tackled by classical analytical methods, which involve analyzing call detail records, mobile recharge information and contract data in an attempt to predict purchase behavior and churn rates. However, these methods often provide results that are "too little, too late," and may be undifferentiated in their targeting.
Often ignored are the importance of the interactions an individual might have with other people, and the inevitable influence that family members, friends and business contacts have on one another. When these social interactions are examined, the result tends to be limited to basic calling-circle analysis, which looks only at the number of different contacts a customer has and whether these people subscribe to the same service.
SNA goes beyond the calling circle. Besides quantifying the strength and influence of each person within a group (see "Social network analysis attribute meters," right), SNA identifies key decision makers within that network. For example, if Suzy is the central communicator within her circle, her choice to use one operator over another may have more impact on everyone around her than if she were an isolated and remote participant. Suzy churning to another service may influence her contacts to churn.
On a deeper level, SNA goes beyond this direct circle to identify the extended network of relationships among customers. In other words, whom do Suzy's friends call?
REDUCE CHURN, INCREASE SALES
Attributes derived from SNA may be used alone or as input to classical predictive models to help improve their accuracy. SNA has already proven its worth in charting relationships in multiple arenas. It has helped identify suspected terrorist and fraud networks and has facilitated an understanding of how organizations interact with one another.
Marketing campaigns aimed at retaining existing subscribers can be enriched by the intelligence gained through SNA capabilities. A growing number of companies already apply SNA to help pinpoint influential customers who are the most likely to drop their service or jump to a competitor. Even a small reduction in churn can mean big savings—the cost of retaining a client is estimated to be only one-fifth that of acquiring one. And these consumers could ultimately help decrease the churn within their own social circles, amounting to even more potential savings.
Additional marketing opportunities are derived from SNA for broadening the scope of existing customers by identifying cross-sell and up-sell opportunities. The social status and position of key individuals in a network can have a profound effect on how their cohorts make decisions. Therefore, the ability to quickly and easily target those influential people can trickle scores of additional profit opportunities across and beyond the entire contact circle.
‘INFLUENTIAL EARLY ADOPTERS’
How quickly new products are accepted in social networks often depends on a relatively few innovative consumers, known in some industries as "influential early adopters." While only about 2% of an organization's customers are estimated to fit into this category, their opinions and purchasing habits typically reach throughout and beyond their calling circles and, in many cases, can be the key to stimulating the market for new products and services. In cases of high-technology products, influential early adopters are often the most technology-oriented in their social networks.
“Social network analysis is fast emerging as an important discipline for predicting and influencing consumer behavior.”
To identify and appropriately target this most valuable—and most elusive—category, SNA analyses and the resulting marketing actions should be run directly in a scalable data warehouse environment. Experts warn, however, that prematurely targeting influential early adopters can prove disastrous for products that have not had their bugs worked out, as the powerful impression these consumers wield can be negative as well as positive.
THE ROLE OF THE EDW
SNA was initially designed to handle networks that number in the thousands rather than in the tens of millions that are commonplace among telcos. Therefore, the biggest challenge in making SNA work is ensuring the appropriate tools are employed to effectively analyze the entire network, no matter how large.
A powerful EDW can ensure valid SNA results even in the largest databases. With SNA, telco leaders can move beyond classical analysis methodologies to more effectively retain customers and easily pinpoint the most influential members of a market for new products and services. This, in the end, will drive businesses to a sustainable goal.
Dr. Judy Bayer is director of Advanced Business Analytics for Teradata in Europe, the Middle East and Africa.
Dr. Edouard Servan-Schreiber, a consultant in predictive modeling, has worked with telco operators, retailers, banks and auto manufacturers.
