The project that received the highest priority was called the happy churn project, as it in fact exhibited one of the telecom company’s biggest problems: an outdated product portfolio, which had not been adjusted along with the development of the mobile market.

As shown in the lead information presented in Exhibit 9.1, about half of all customers who left the company fell in the category of happy churn. These were customers who were happy enough with the company but merely had received a better offer. Should such churners be maintained, it would be a question of contacting them at the right time with the right offer. This kind of knowledge can be generated through a churn prediction model, which is a statistical model that, based on the customers who left the company during the last period, can create a profile of them based on data warehouse information. We could then interpret this profile as an independent analysis that gives input about why we were losing in the market. Furthermore, we could assume that next month will be like the last, which means that the profile can be used for scoring all existing customers. Based on this, we knew on a one-to-one basis whom to contact with what retention offer based on how likely the individual customer was to churn and why.

Exhibit 9.3 shows a tree structure that groups customers by their churn indicators. Based on this model, we could describe the overall approach the company should follow. The figures in the model ...

Get Business Analytics for Sales and Marketing Managers: How to Compete in the Information Age now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.