Hello, I am working on Customer churn analysis. I need to identify what data I can use from operators' web portal regarding that customer. I mean I have data regarding the web pages customers have visited, how many times, for how long etc. Can someone help me to identify what all such metrics can be used in analysis. I will be using R for my analysis. Thanks is advance.
Hi, I believe your description is incomplete and it is quite difficult to come up with a spot on answer. Nevertheless, assuming that you are doing this analysis on behalf of an operator that obviously collects this data from its network, you will not be able to predict the churn, unless you have analysed these statistics over a longer period of time and have for instance details of how many of the customers actually left after they visited competitors websites and how long on average did it take them to leave since first explored this possibility. In R you may use the hypothesis testing (p value method). Since I cannot quite get what is it what you are trying to achieve is impossible to assist you further. If for instance you are trying to prevent the churn think this way: People usually search for better deals, phones, plans, etc, right before their previous locked in contracts expire(one month or so) so if I would like to prevent the churn, I would start by analyzing this particular segment and their behaviour. I hope these ideas were helpful to some extent.
I am interested in this field. Frankly, I think we could not identify what methods can be used for your project. Firstly, the metrics are closely related to your research object. It is hard to define specific indicators regardless of the research issues. Secondly, the proper methods cannot be found before we analyze the data set. I would like to help you if you prefer to send a detailed description.
- Badges
- Science topic
- Similar topics
- Engineering
- Telecommunications Engineering