I have dataset representing to specific region of the country. Its purchase data which is how much costumer is placing with associated dates of order, amount etc.
I want to do churn modeling for this dataset. However, I have an issue,
- I have initial customer which goes back to Dec 2018.
- Then we have another business acquired and this data starts from October 2021 which makes 70% of our customer base.
- I can find churn window based on repeat transactions
Now I am confused, if I can use all data for churn modeling?
- How am I going to justify given that those are customers joined way later than other initial base?
- Is there any other way around so I can model for all dat instead only 30% of customers data?
- Is there any rule base approach I can use here?
I would appreciate your advice and suggestions.
Thanks in advance