Churn Modelling – Notes from the Digital CRM Video Hangout
Raja Mitra has recently organised a very informative digital CRM web meeting where Dr Abhijit Sanyal delivered a lecture on predictive analytics and usage of predictive analytics of big data on the video. He described three case studies, one for banks how customer satisfaction and employee satisfaction were linked. The second case was of churn propensity modelling applied in Telecom sector (how to reduce cost of targetting using Churn propensity modelling).
Thoughts on Churn Modelling
Thinking about the churn modelling from Abhijit’s presentation, I was wondering if we can set up churn models for student churn through the university systems and patient churns through health systems, and then bring together this model through the design thinking aspects of doing a student or patient journey modelling. The churn modelling aspect was particularly informative.
“Churns”, or understanding why “clients”, after an initial engagement with the unit do not come back, or “fail to survive through the process” is particularly interesting in case of college admission units. For example, each year, the School of Health Sciences receives queries from hndreds of students; it’d be insightful to test what proportion of that student query actually end up getting admitted. Also, once being admitted, a certain percentage of the student move away from the university and move away. What is the attrition rate of differnt programmes and here too, churn modelling will have a role to play.
The worth of churn modelling in the educational sector such as a university admissions department will be similar to the Internet companies who will like to retain their customers, as holding on to established customer bases and trying to understand their behaviour is much more beneficial than acquiring new customers.
A similar issue works with the health services community. We need to know what is the “churn” of patients or clients who initially engage with the preventive or health promotive activities and then move away.
An Initial Ideas for churn modelling in Education and Health Sector — Methods
//To be filled in