2
$\begingroup$

I have longitudinal transaction data of a retail store where each row is a transaction done by an individual. I would like to perform a survival analysis to analyse how long a customer will transact before churning. For CoxPH model, it requires tenure and churn variable among other variables. I am not sure what is the correct approach to prepare these variables for this analysis. How to label whether a person has churned other than having a threshold of 45days or 2 months to decide it. Also, what is the correct way to represent tenure.

Below is the sample of data

Id.    Visit_date.       Amount.   Tenure     Churn     Age     Income
1.     04/03/2020        500        ?           ?       40      56K
1.     05/03/2020        300        ?           ?       32      60K
1.     05/23/2020        800        ?           ?       28      90K
1.     07/04/2020        700        ?           ?       40      56K
2.     02/03/2020        500        ?           ?       43      50K
2.     05/12/2020        300        ?           ?       60      90K
3.     03/23/2020        800        ?           ?       18      80K
4.     07/04/2020        700        ?           ?       20      40K

```
Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

It is your decision on what you define as churn. This is something to discuss with business people who are going to use those results. The definitions will vary on case-by-case basis. Usually what is used is some threshold that is meaningful from business perspective.

Improve this answer
$\endgroup$
1
  • $\begingroup$ thanks for the response. Also, suppose a threshold is used to define the Churn, then how should i setup tenure variable in case of this longitudinal data considering the above sample data where user_id 1 has made four transactions. $\endgroup$
    – krishna koti
    Commented Jul 31, 2020 at 21:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.