# Bayesian Model For Churn

I need to evaluate how long a customer stays with the company given a retention offer she accepted $r\in\{r_1,\dots,r_k\}$

I'd like to use Bayesian inference for modelling churn.

What prior distribution should I choose to model churn rate ?

• What is the likelihood you intend to use? – ilanman Dec 16 '15 at 1:52
• I think that exponential makes the best sense – Uri Goren Dec 16 '15 at 10:17