Abstract:Currently, almost all direct marketing activities take place virtually rather than in person, weakening interpersonal skills at an alarming pace. Furthermore, businesses have been striving to sense and foster the tendency of their clients to accept a marketing offer. The digital transformation and the increased virtual presence forced firms to seek novel marketing research approaches. This research aims at leveraging the power of telemarketing data in modeling the willingness of clients to make a term deposit and finding the most significant characteristics of the clients. Real-world data from a Portuguese bank and national socio-economic metrics are used to model the telemarketing decision-making process. This research makes two key contributions. First, propose a novel genetic algorithm-based classifier to select the best discriminating features and tune classifier parameters simultaneously. Second, build an explainable prediction model. The best-generated classification models were intensively validated using 50 times repeated 10-fold stratified cross-validation and the selected features have been analyzed. The models significantly outperform the related works in terms of class of interest accuracy, they attained an average of 89.07\% and 0.059 in terms of geometric mean and type I error respectively. The model is expected to maximize the potential profit margin at the least possible cost and provide more insights to support marketing decision-making.

What problem does this paper attempt to address?

The main problem this paper attempts to address is: In the context of the current digital transformation and the increasing number of virtual marketing activities, companies are increasingly relying on remote methods such as telemarketing to understand and predict customers' purchase intentions. However, due to data imbalance (i.e., the number of customers willing to accept marketing proposals is much smaller than those who are not), data missing, and the influence of external factors (such as market dynamics, economic stability, etc.), building effective predictive models becomes very challenging. Moreover, traditional marketing analysis methods often fail to comprehensively handle these complex data characteristics. To address these challenges, this study proposes a new method based on Genetic Algorithm (GA), aiming to simultaneously select the best distinguishing features and optimize classifier parameters. Specifically, the goals of this study include: 1. **Develop a high-performance predictive model**: Capable of accurately predicting whether customers are willing to accept the offer of a term deposit. 2. **Identify the most significant customer features**: Find out the key factors that influence customers' acceptance of marketing proposals. 3. **Maximize potential profit**: Achieve the maximum profit margin at the lowest cost. 4. **Support marketing decisions**: Provide more insights to support the company's marketing decision-making process. Using real data from a Portuguese bank and national socio-economic indicators, this study constructs an interpretable predictive model and validates the generated optimal classification model through 50 repetitions of 10-fold stratified cross-validation. The results show that the model significantly outperforms related work in terms of accuracy for the target category, achieving a geometric mean of 89.07% and a Type I error rate of 0.059. This indicates that the method not only improves predictive performance but also provides more valuable business insights.

Modeling the Telemarketing Process using Genetic Algorithms and Extreme Boosting: Feature Selection and Cost-Sensitive Analytical Approach

Marketing Intelligence in the Era of Big Data

Data Pre-Processing Algorithm for Neural Network Binary Classification Model in Bank Tele-Marketing

A Deep Learning Approach To Bank Direct Marketing

A dynamic ensemble selection method for bank telemarketing sales prediction

GATeS: A Hybrid Algorithm Based on Genetic Algorithm and Tabu Search for the Direct Marketing Problem

Modified Genetic Algorithm for Feature Selection and Hyper Parameter Optimization: Case of XGBoost in Spam Prediction

Imbalanced customer classification for bank direct marketing

A two-step system for direct bank telemarketing outcome classification

Dynamic Marketing Resource Allocation with Two-Stage Decisions

Optimizing Churn Identification in Telecommunications Using Natural Language Processing and XG Boost Machine Learning Paradigm

Dynamically Tie the Right Offer to the Right Customer in Telecommunications Industry

Modeling of an artificial intelligence based enterprise callbot with natural language processing and machine learning algorithms

Bayesian Regression for Predicting Subscription to Bank Term Deposits in Direct Marketing Campaigns

Binary Classification of Customer’s Online Purchasing Behavior Using Machine Learning

Data-driven agent-based exploration of customer behavior

Multiple Resource Allocation for Precision Marketing

Customer Profiling, Segmentation, and Sales Prediction using AI in Direct Marketing

Case Study of Model Selection on Customer Information Task Based on Machine Learning Algorithms

Machine Learning Application in Car Insurance Direct Marketing

Intelligent Classification-Based Methods in Customer Profitability Modeling