Question

How does one predict future purchase of the customers who have purchased our product or service recently for first time with the help of modeling?

Suppose 4000 customers purchased iPhone Handset first time in Delhi city during December 2020. We want to project year-wise handset purchase behavior for next 6-7 years and also find how many will repurchase iPhone year-wise based on some historical data.

Past data indicates that people in Delhi usually purchase a new mobile every year ( 25% chance) using facilities like exchange or sometimes gift older handsets. There is 60 % chance of buying a new phone at the end of two years and mostly people using iPhones don't carry older devices more than three years.

The brand loyalty of buyers towards iPhone grows over time. Therefore, customers who purchase a new device after one year after purchasing iPhone, there is a 60% chance they will buy iPhone again and if they buy after 2 years or 3 years, the chances are 70% and 80%.

Some assumptions:

Mobile Users buying high end mobile devices only buy new phones only during Christmas. Same customer may sometimes buy new phone at the end of one year and may be next time after two years or three years. One user doesn't keep more than one phone.

If a user switches brand, he/she doesn't purchase same brand again.