Telecom Operator Case Study – A comparison between traditional methods and Xtract’s community marketing intelligence approach.
Customer
A UK based telecom company that offers broadband, fixed line and prepaid mobile connections.
Challenge
Our client was faced with a significant challenge to reduce high rate of churning customers. Xtract knows that a subscriber’s immediate neighbours in their social network and their actions have a real impact on churn.
Field Trial of the Automated Solution
To prove the positive effect of Xtract’s Social Links, we tested it in two separate campaigns for the same service. The product in focus was a fixed line service where the subscribers can call for free amongst the network and at a reasonable price outside the network, against a monthly fee.
The two campaigns were:
A “traditional” churn management campaign, where targeting is typically based on the analysis of the subscriber’s individual behaviour
A campaign where the churn propensity scoring was also influenced by the social network. By appropriate targeting of highly influential, high propensity churners in the social network based campaign, one could reduce churn not only among those targets, but also in their communities.
Four groups were chosen as targets:
Two for the traditional and the other for social networks based campaign
Two groups as controls for the target groups respectively, and the churn in all four groups was measured over a three month period.
Results
The results were clearly in favour of Xtract’s social network based campaigns: the churn was reduced by 26% compared to control group, versus reduction of 19% in the conventionally targeted group – a 37% relative improvement!
Reducing churn has a major effect on the operator’s bottom line results, so the impact of a successful churn campaign can be huge.
Using CDR and network analysis in addition to demographic data produces in this case a 21% better accuracy in predicting churn than a corresponding regression model based on demographic information only.
