With penetration of mobile phone usage approaching and even exceeding 100 percent in many countries, mobile operators are becoming more and more dependent on data generated by their legacy transactional systems to innovate new products and implement customer retention campaigns.

Customer churn has become one of the major indicators that mobile operators keep their eyes on, when planning and reporting market share dynamics. And although most of the operators have their strategies and a complex of measures to reduce churn and increase customer loyalty, industry wide researches, including the one implemented by Enders Analysis on UK mobile market in 2010, demonstrate the persistence of customer churn management challenges and its growing concern.

Since the recognition of customer churn as one of the major drivers of success in mobile service businesses, numerous researches were made and studies published on usage of various statistical and data mining techniques to model and predict customers prone to churning.

Most of these studies however concentrated on the effectiveness of certain techniques to “calculate” which customer will most probably churn, so as to find the main target for a new product or service proposition. Academic researches covering customer churn in telecommunications provide little evidence on development of special products and price offers related to variety of very specific groups of churners. In most of the cases the churners are just one group to target the new offer.

There is strong evidence that many managers responsible for churn reduction coupled the underlying technology for service provision with statistical techniques to make their best estimates on the subject of “Who will most probably churn”. The input and output of data mining tools where in their majority limited with the “technological centricity” that most of telecommunications and other infrastructure technology businesses exhibit.

A number of academic and empirical studies argue that these companies are “businesses organised around technology” and not “Technologies designed to make the business more efficient”. One of the key manifestations of the “technological centricity” is that many mobile operators are making their strategic decisions under constant pressure of recovery from sunk coasts and not guided by short term market consideration.

In terms of data mining for churn prediction, this approach guides the companies first of all to utilise the existing transactional, billing and contract management technology, which, in the first place, was not designed with churn prediction consideration in mind.

This is why the rich information that customers leave with us throughout all touch points travels through the wealth of statistical modelling tools and techniques, just to arrive with true and false alarms limited to divide between churners and non-churners.

However the more data customers leave with mobile operators in form of reached phone numbers, SMS and data usage quantities and accessed web contents and services through a variety of applications, the more opportunities they raise and ask for innovative services at more affordable price.

Today, customers expect from their mobile operators rich context aware propositions similar to the ones they receive from various web services utilising all the wealth of behavioural data. The customers want to enjoy the same relevance of offers as they see with amazon.com, offering relevant books bought by other customers who bought a particular book.

Mobile customers are demanding their offers to be made at the most relevant time, at the most meaningful price in terms of value proposition for a tailored service-product. The customer wants these offers to change with seasonal and their own age cycles, to travel with them, when they are travelling and to go for vacation with them when they are away from their everyday business activity. The customers want the offer to acknowledge their relationships with local and international friends and partners.

These are all contextual marketing considerations that can be extracted from the data generated by legacy platform.

By recognising non-relevance and resulting low response rates of offers, telecoms make an important move towards new promises that customer centric technology developers bring to get mobile operators out of the trap called “Technological Centricity”.