Bibliographic citations
Mori, E., (2022). Optimización de las campañas de marketing en la industria financiera: un enfoque basado en machine learning [Tesis, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/26917
Mori, E., Optimización de las campañas de marketing en la industria financiera: un enfoque basado en machine learning [Tesis]. PE: Universidad Nacional de Ingeniería; 2022. http://hdl.handle.net/20.500.14076/26917
@mastersthesis{renati/712794,
title = "Optimización de las campañas de marketing en la industria financiera: un enfoque basado en machine learning",
author = "Mori Orrillo, Edmundo de Elvira",
publisher = "Universidad Nacional de Ingeniería",
year = "2022"
}
Marketing campaigns implemented by financial organizations are focused on the product. The companies in the system offer an average of 17 different products between assets, liabilities and services; obtaining that the effectiveness and profitability of these campaigns is not optimal. Therefore, to increase these indicators, it is necessary to build an analytical model called “Next Best Offer“ (NBO) that allows defining the set of products to be offered, prioritized according to their propensity and profitability. To develop this model, first of all, the expected profitability is calculated using predictive models that use the following algorithms: Linear Regression, Random Forests, Regularized Linear Regression, Regression with Support Vectors, etc; secondly, we calculate the acquisition probabilities of each product using classification models that use the following algorithms: Logistic Regression, Decision Trees, Support Vector Machines, among others. Different analytical models are used to calculate acquisition probabilities, due to this to compare them with each other, it is necessary to standardize them using a single model. This standardization can be done using assembled models that combine different classification algorithms or using a model based on artificial intelligence called Neural Networks. The comparison of the prediction capacity of each method indicates that the assembled model obtains the best standardization results, this capacity was calculated using the indicator Area Under the Curve (AUC, acronym in the English language). The standardized probability and profitability will be the variables used to build the prioritization function, output of the NBO model, which allows ordering financial products according to their degree of acceptability. The NBO model allows the Business Intelligence area to build marketing strategies customer vision with greater effectiveness and profitability; This affirmation is demonstrated by analyzing the results of the commercial campaigns of the first 6 months of the year 2021. These campaigns were carried out using marketing strategies product vision, in their construction, empirical rules called expert rules were used, and comparing them with the results of the commercial campaigns of the last 6 months of the same year (strategies built using the NBO model). The comparison obtained as a result that the effectiveness and profitability of each campaign increased substantially in the last 6 months.
This item is licensed under a Creative Commons License