Hi
i have a dataset of monthly invoices over 5 years. Some customers only have 3 years worth of data
invoice value is based on transactiin amt x fee
I also have another dataset for the same customers / timeframe giving the number of API calls
I want to forecast the probability of a customer leaving
how would I go about this?
It is difficult to answer the question without data.
One thing, is to try and chart the two values over time (transaction amount x fee vs. number of API calls) to see whether they are correlated.
To figure out PROBABILITY, maybe look at the data of other customers “leaving” versus their behavior over time. This can give you a clue regarding probability. If you compare the two populations (leaving versus staying), you may be able to forecast the probability.
If you want to hide the customers’ names and any other information – you can sent me sample data ([email protected]), this way I can get an idea and try to help
