Customer churn data analysis using data mining

Introduction. Customer churn is one of the key challenges in the telecommunications sector, significantly impacting companies' financial performance. Losing customers leads to substantial economic losses, as acquiring new users is typically more expensive than retaining existing ones. This creates a need for the development of effective analytical approaches to predict customer churn and identify the factors influencing it. Objective. This study aims to assess the effectiveness of survival analysis methods in predicting customer churn in the telecommunications sector. The primary objective is to identify key factors affecting churn rates and evaluate the impact of targeted retention strategies on customer behaviour. Methodology. The study employs survival analysis methods to evaluate customer churn dynamics. The Kaplan-Meier estimator is used to analyse the temporal patterns of customer retention, identifying high-risk periods. The Cox proportional hazards model is applied to determine the influence of various factors on churn probability. The dataset consists of real-world telecommunications customer data, including demographic characteristics, contract details, and financial indicators. Preprocessing steps include data cleaning, feature selection, and transformation of categorical variables into numerical format. Model performance is evaluated using metrics such as the Concordance Index (C-index) and Brier Score. Finally, A/B testing is conducted to assess the effectiveness of retention strategies in reducing customer churn. Results. The findings indicate that the highest risk of churn occurs within the first 20 months after subscription activation. It was determined that long-term contracts and high-quality service reduce the likelihood of churn, whereas high monthly costs and certain payment methods increase it. The developed predictive models achieved an accuracy level of 85–87%. Implemented retention strategies within A/B testing resulted in an increase in the retention rate from 74% to 91%. The proposed methods can be applied by telecommunications companies, financial institutions, and other businesses operating in the subscription-based service industry to enhance loyalty programs and reduce customer churn.

With the rising growth of the telecommunication industry, the customer churn problem has grown in significance as well. One of the most critical challenges in the data and voice telecommunication service industry is retaining customers, thus reducing customer churn by increasing customer satisfaction. Telecom companies have depended on historical customer data to measure customer churn. However, historical data does not reveal current customer satisfaction or future likeliness to switch between telecom companies. The related research reveals that many studies have focused on developing churner prediction models based on historical data. These models face delay issues and lack timelines for targeting customers in real-time. In addition, these models lack the ability to tap into Arabic language social media for real-time analysis. As a result, the design of a customer churn model based on real-time analytics is needed. Therefore, this study offers a new approach to using social media mining to predict customer churn in the telecommunication field. This represents the first work using Arabic Twitter mining to predict churn in Saudi Telecom companies. The newly proposed method proved its efficiency based on various standard metrics and based on a comparison with the ground-truth actual outcomes provided by a telecom company.

Customer churn prediction plays a vital role in modern business, accurately influencing strategic and operational decisions that influence customer loyalty to a service. Customer churn focuses on customer retention being more profitable than attracting new customers because long-term customers provide lower profits and costs while losing customers increases the costs and need to attract new customers. However, customer churn still occurs frequently and cannot be predicted. If customer churn is left unchecked, it will endanger the company or banking industry because it can cause loss of income, damage reputation, and decrease market share. Random Forest, a data mining technique, was used in this research because of its ability to predict and handle many variables. This research aims to predict customer churn using the Random Forest method with datasets from Europe, especially France, Spain, and Germany, hoping to benefit the banking industry by identifying customers at high risk of abandoning services. This research is expected to benefit business people from customer churn predictions. Especially in the banking industry, it can help identify customers at high risk of abandoning service. Thus, companies can take appropriate steps to retain these customers, increase customer retention, strengthen customer loyalty and optimize their business performance. The results of this research are an accurate system for predicting customer churn in the future. The research obtained accuracy results of 87% in predicting customer churn using accuracy testing in the form of a confusion matrix.

Customer churn indicates percentage of customers that are no longer using a company service in a certain time. Thus, it leads to decrease of the company profit. PT. XYZ, one of Telecommunication Company in Indonesia, approximately has a churn rate of 18%, therefore the company needs to focus its improvement in order to lower the churn rate. This study aims to build a classification model based on optimized deep learning ANN algorithm in order to predict customer churn rates. Deep-learning ANN has an advantage of its flexible characteristic. The result show that the optimized ANN model has an accuracy value of 76,35%. The model is built with an epoch parameter of 30, hidden layer =50 with tanh as the activation function. The contract type, type of service, and IPTV are the three most influential variables in customer churn at PT. XYZ. The prediction results in the optimized deep learning-ANN model indicate that there is 2567 customers tend to be churn customers and 4386 customers tend to be loyal customers. The churn percentage in this model is 36.4%. There is an increase of 11% compared to the previous year.

Introduction: The Churn means Attrition. In the world of competition and customer needs, the superior quality service, to predict customer churn, has become a pressing issue for the HRMS Vendors or service providers. Machine Learning is capable to solve this challenging problem. The customer churn can be predicted, and proactively, HRMS vendors can apply a strategy to retain customers and manage the customer relationship. Objectives: The objective of this paper is to meticulously analyse and summarize the previous studies conducted for customer churn prediction. Furthermore, this paper will serve as a valuable resource for future research in the context of HRMS vendors or service providers. Methods: This paper focuses on machine learning algorithms like Decision Tree, Random Forest, Support Vector Machine, Gradient Boosting, Neural Network, etc. The paper highlights the accuracy of the outperformed models. This paper includes the previous studies between 2021 and 2025. In total, 30 papers were reviewed meticulously. To provide more clarity, all the previous studies are summarized in a table format. The paper was searched on Google with the keywords ‘Customer Churn’ AND ‘Machine Learning’. The Duplicates were removed, and the previous studies were analyzed. Results: The findings showcase that various machine learning models like RF, DT, Logistic Regression, Gradient Boosting are applied to predict churn. Furthermore, by this analysis it is clearly stated that hybrid or ensemble models work best for customer churn model prediction. This study guides on selecting the best model as per the relevant industry. Conclusions: This comprehensive survey has meticulously analysed the implementation of machine learning algorithms to predict customer churn and highlighted its importance to fully transform the business to build client retention strategies. In this paper, different machine models are examined, and for each previous study, the machine learning models used & their results are discussed.

Customer churn—when a customer stop buying a company's products or using its services—has a bad influence on the revenues. Some industries have churn rates of around 20-40% each year. The paper presents a strong model of estimating the customer turnover based on the Kaggle Telecom turnover Dataset. The proposed technique consists of extensive data preprocessing that gets rid of irrelevant features, fills in missing values, detects outliers, and corrects class imbalance with the help of the SMOTE-ENN method. Pearson correlation analysis and visualizations are very helpful in understanding the relationships between features and in reducing dimensions while retaining important information. On the other hand, Principal Component Analysis (PCA) eliminates some of the data’s dimensions but retains the significant information. The accuracy (ACC), precision (PRE), recall (REC), F1-score (F1), ROC curve, and confusion matrix are some of the performance metrics used to train and evaluate an XGBoost classifier. The accuracy score is 98.42. The experimental findings show that the suggested model outperforms the existing machine learning approaches and has extremely good prediction performance. The study's findings validate the model's use in the creation of proactive customer retention strategies within the telecom industry by demonstrating its capacity to properly anticipate customer attrition.

Customer churn exhibits significant variation influenced by both individual characteristics and industry sector. This study analyzes two large-scale datasets comprising thousands of observations, employing logistic regression and survival analysis to investigate churn determinants. Key findings indicate that the overall churn rate in the finance sector (20.37%) is significantly lower (by 6.17 percentage points) than in the telecommunications industry (26.54%)(p<0.05). Furthermore, significant gender disparities exist: males demonstrate lower churn rates than females, with a non-significant difference of 4% (p=0.44) in telecommunications but a substantial 36% reduction in the banking sector (p<0.005). Churn propensity is positively associated with age, with older customers exhibiting significantly higher churn rates than younger cohorts73% higher in telecommunications (p<0.05) and 23% higher in banking (p=0.03). Survival analysis revealed distinct retention patterns across customer nationalities: the German customer group exhibited the highest churn risk (fastest decline in survival probability), followed by the Spanish group (moderate risk), while the French group demonstrated the most favorable retention performance (smoothest survival function decay; p<0.05 for group comparisons). These results have significant implications for the development of targeted customer retention strategies. This study makes a substantive contribution to the customer churn literature by transcending the prevailing single-industry paradigm.

Improving the diagnosis and prediction of customer churn: A heterogeneous hazard modeling approach

Customer loyalty is crucial for the long-term success of service businesses, particularly in the online transportation industry. It involves customers consistently choosing to use a company's services or products. This research seeks to examine how both pricing and service quality can influence customer loyalty within the online transportation industry by looking at factors like customer trust and satisfaction. The study utilizes a quantitative methodology with a descriptive design. The population in this study were users of K-JEK services whose number was unknown but had used K-Jek transportation services, while the number of samples was 273.25 and rounded up to 280 samples of K-Jek transportation service users. The method of collecting data involved using a survey with a Likert scale. The analysis of the data employed SEM (Structural Equation Modeling) or Structural Equation Model through the SmartPLS software. The outcomes show that Service Quality and Price impact Customer Loyalty through Customer Satisfaction and Trust. High-quality service enhances satisfaction and trust, leading to greater loyalty. Competitive pricing also improves customer perception, fostering stronger trust and retention. The findings confirm that Customer Satisfaction and Customer Trust play a crucial mediating role in strengthening these relationships. This study suggests that online transportation providers should prioritize service quality and fair pricing strategies to enhance customer loyalty. Maintaining high service standards and transparent pricing can improve customer satisfaction and trust, ultimately securing long-term success.

Customer churn is a key issue in the telecom industry and has a significant impact on the revenue of related companies. For telecom companies, establishing an accurate customer churn prediction model is crucial for retaining existing customers. Therefore, it is necessary to develop reliable prediction models to effectively manage and reduce churn rates. This study explores how to use machine learning technology to predict customer churn based on various customer characteristics and usage patterns, and its application in accurately predicting customer churn based on demographic data, service usage behavior, and billing information. Through systematic analysis of model performance, the advantages and disadvantages of each method are highlighted, with particular emphasis on the superior predictive power of the integrated method when dealing with imbalanced data sets. This proves the potential of these methods in accurately predicting customer churn behavior, thereby providing relevant companies with methods to retain customers. In addition, the study explored the economic implications of the accuracy of attrition predictions and advocated the use of lower-cost methods to ensure the rational allocation of resources.

Customer churn is the loss of business clients to a competitor. Since keeping current clients is more economical than finding new ones, customer retention measures such as churn detection are now essential aspects of modern banking strategy. However, many existing studies rely heavily on conventional machine learning approaches such as Support Vector Machines, Logistic Regression , Random Forest, etc., often neglecting the deeper learning capabilities of neural networks. Also, the repeated use of the same small dataset by the banking studies may limit the improvement of the models’ generalisation. To address these gaps, this study presents a method that integrates deep learning for customer churn detection and a soft and hard voting classifier ensemble embedded with the best performing models over the years for results comparison, supported by a synthetic data augmentation method for model improvement. The study utilised a secondary banking churn dataset from Kaggle, which contained 10,000 unique customer records. To address the dataset limitations, a Conditional Tabular Generative Adversarial Network (CTGAN) model was used to generate an additional 10,000 records, expanding the dataset used for the study to 20,000 rows. Data preprocessing steps were done before training, including oversampling using Synthetic Minority Oversampling Technique (SMOTE). Model development and analysis processes were implemented using Python programming language with prominent libraries and frameworks on Google Colab. In this study, a Feedforward neural network and a soft and hard voting classifier were developed. The voting classifier ensembles integrated three prominent classifiers: Random Forest, XGBoost, and Logistic Regression. The performances were evaluated using Accuracy, F1 Score, and Area Under ROC Curve as metrics. Results show that while the Feedforward Neural Network achieved strong predictive performance with an accuracy of 88.23%, an F1 Score of 87.83% and an AUC of 94.73%, the ensemble approaches performed slightly better as the soft voting classifier delivered the best results, obtaining an accuracy of 89.46%, F1 Score of 88.92% and AUC of 95.40% showing the advantage of combining multiple models to leverage complementary strengths. After comparison with past studies, the proposed models did not surpass the very best outcomes. However, they remain highly competitive, achieving performance levels that are on par with or exceed many earlier works. The contribution of this work is to show how synthetic data augmentation, enhanced preprocessing, deep learning techniques, and machine learning ensembles can improve churn detection in banking studies. Banking institutions can utilise the results from this study to accurately detect churn, supporting proactive customer retention strategies, targeted marketing, and personalised financial services, thereby reducing revenue losses.

This study explores customer behavior towards secondary banking products, including credit cards, insurance, and wealth management services, among clients of commercial banks. It identifies key factors influencing adoption decisions, such as product awareness, perceived value, and trust in the bank, while also examining barriers to uptake. The findings provide insights for banks to enhance cross-selling strategies, improve customer satisfaction, and increase retention through tailored product offerings. This case study investigates the impact of credit cards, insurance, and wealth management products on customer retention within commercial banks. Through a comprehensive analysis of customer behavior and preferences, the study identifies key factors that influence retention rates. This descriptive study, aimed to analyze customer behavior towards secondary banking products such as credit cards, insurance, and wealth management services. Utilizing a simple random sampling method, a sample of 330 customers was surveyed using a questionnaire as the primary data collection tool. The findings revealed that these secondary products have a minimal impact on customer retention, indicating the need for banks to develop more attractive schemes to enhance customer loyalty and retention. Furthermore, effective wealth management strategies tailored to individual customer needs contribute significantly to long-term retention. The study concludes by highlighting the importance of personalized financial services in fostering strong relationships and reducing customer attrition in the competitive banking sector.

Nowadays, telecommunication markets are becoming more and more competitive, and customer churn is becoming more and more serious. In the tough competitive mobile market, Customer Churn Management is becoming more and more critical. In developing countries, most customers switch service providers because of good promotional incentives and lower monthly costs offered by competitive service providers. How to predict customer churn quickly and accurately becomes very important. In this paper, the researchers successfully analyzed the customer churn using big data feature analysis and multi-feature analysis. User data were modeled by XGBoost algorithm. The model is optimized repeatedly with GridSearchCV as a parameter tool. The accuracy of the model on the test set is 85.1%. The researchers predicted about 11000 customer lists per month that may be about to churn. Using K-means clustering method, 11000 churn target customers per month were classified into three categories and telecom companies are suggested to take some solutions which are found by feature analysis to retain customers. This big data analysis can be used to retain customers for the telecom industry.

A service provider is a business that provides services or the expertise of an individual in a certain sector. A service provider’s customer flow could be very dynamic, with both new and churning customers. For the purpose of minimizing the number of churning customers, the company should perform a customer churn analysis. Customer churn analysis is the process of identifying a pattern or trend in churning customers. In order to classify and predict churning customers, machine learning techniques are required to build the classifier model. This paper will use the Support Vector Machine (SVM), Light Gradient Boosting Machine (LightGBM), and hybrid Adaptive Boosting-SVM (AdaBoost-SVM) model. The hybrid AdaBoost-SVM model is a boosting model which uses SVM as its basis classifier instead of a decision tree. The models will be implemented using airlines and telecommunication customers churn data. The usage of oversampling technique is required to balance the number of observations in both classes of training data. Furthermore, a model comparison will be conducted using the F1-Score and the AUC score as the evaluation metric. The analysis shows that LightGBM performs the best result in both dataset with the highest F1-Score and the shortest computational time. In addition, the boosting model AdaBoost-SVM has a better performance than the SVM model due to the boosting algorithm which always minimizes the model error in each iteration. Despite having a better result, AdaBoost-SVM performs in the longest computational time, making it computationally expensive for large datasets. Additionally, the imbalanced nature of the datasets presents challenges in model performance, requiring the application of oversampling techniques to mitigate bias towards the majority class. In conclusion, LightGBM is the best model to classify churning customers based on the higher F1-Score, AUC score, and the shortest computational time.

Customer churn describes terminating a relationship with a business or reducing customer engagement over a specific period. Two main business marketing strategies play vital roles to increase market share dollar-value: gaining new and preserving existing customers. Customer acquisition cost can be five to six times that for customer retention, hence investing in customers with churn risk is smart. Causal analysis of the churn model can predict whether a customer will churn in the foreseeable future and assist enterprises to identify effects and possible causes for churn and subsequently use that knowledge to apply tailored incentives. This paper proposes a framework using a deep feedforward neural network for classification accompanied by a sequential pattern mining method on high-dimensional sparse data. We also propose a causal Bayesian network to predict cause probabilities that lead to customer churn. Evaluation metrics on test data confirm the XGBoost and our deep learning model outperformed previous techniques. Experimental analysis confirms that some independent causal variables representing the level of super guarantee contribution, account growth, and customer tenure were identified as confounding factors for customer churn with a high degree of belief. This paper provides a real-world customer churn analysis from current status inference to future directions in local superannuation funds.