Probabilistic models of banking products and their use in marketing optimization

The main objective of the thesis is to use the results of probabilistic models in marketing optimization and to make the process of approaching clients to sell banking products more efficient. Probabilistic models were developed several years ago in connection with other projects where it was necessary to know the probabilities of sales of individual products. We are now encountering other uses that can be applied in marketing optimization. The purpose of marketing optimisation is to find the solution that will maximize the bank's profit in terms of expected revenue each month. The expected revenue includes the operating costs of reaching clients, which complicates the calculation. The solution to such problem is to find optimal combination between the product and the communication channel for each client. This leads to the need to be considered a three-dimensional assignment problem in which capacity constraints for some communication channels need to be taken into account. Ultimately, the whole problem is converted into a two-dimensional problem, with the first step being the calculation of the probability of buying the product, followed by the calculation of the expected revenue for each client and each combination of product and communication channel, and finally the solution of a mathematical model that considers the capacity constraints.