A Conceptual Framework of the Factors Influencing Customer Churn in Telecommunication Companies

A Conceptual Framework of the Factors Influencing Customer Churn in Telecommunication Companies

Sibusisiwe Dube, Newton Nyamugure

Department of Informatics and Analytics National University of Science and Technology Bulawayo, Zimbabwe

Sinokubekezela P. Dube

School of Engineering The University of Lusaka Lusaka, Zambia

Abstract – AbstractMany sectors, the telecommunication industry in particular, have experienced a massive customer churn due to reasons including dissatisfaction with the service provision. An increase in customer churn has subsequently resulted in the reduction in revenue collection within the telecommunications industry. Attempts to retain customers have not been successful. Moreover, there is limited literature that discusses the factors influencing the increase in customer churn. Adding to the problem is a lack of studies that provide a conceptual framework of factors useful for predicting customer churn. An understanding of these factors is important for retaining the customers, which translates to an increase in the revenue collection within the telecommunications industry. It was with this in mind that this study presents a conceptual framework of factors for predicting customer churn. This study was a systematic literature review, which followed the guidelines of the PRISMA model. Included in the study were 28 peer- reviewed articles published in the IEEE and Springer Nature Link databases, written in English and published between 2023 and 2025. The results show several factors, including customer demographics. These findings also identified machine learning models as the integral models for predicting customer churn in the telecommunications industry.

Keywords – customer; churn; prediction; telecommunication; company

1. INTRODUCTION

   Customer churn has become a major concern in the telecommunications industry [28], which occurs when customers quit doing business with a company [17], [31]. The dynamic nature of the telecommunications industry and varying customer preferences are threatening an attempt to retain customers [25], [30]. It is these industry conditions that are accelerating the need to understand the causes of customer churn and to develop strategies to mitigate it [24]. Furthermore, the cost of acquiring new customers is very high due to the increase in competition in the telecommunications industry [26], [29]; hence, the urgent need to retain existing customers [33], [34]. Prediction of customer churn in these companies can be an effective strategy for retaining targeted customers. Moreover, customer retention has become a critical factor for sustaining profitable businesses [27], [29], [32] such as in telecommunications. Identifying and retaining such customers could guarantee consistent income, reduce promotional expenditure, and promote sustained expansion, thereby increasing the companys revenue while decreasing the

   customer churn rate. Literature shows that while strategies have been implemented for customer retention, there is still an increase in the number of customers who leave the telecommunications industry, an indication that the existing customer retention programs have not been successful [18]. Nonetheless, the increasing growth in global e-commerce greatly demands effective customer retention strategies. Furthermore, it has been noted that even a slight decrease in customer retention can have detrimental effects on the companys revenue. It is therefore important to identify and implement relevant customer retention strategies if the projected 2027 revenue of US$8 trillion is to be achieved, thereby doubling the 2023 revenue of US$5.8 trillion in the same telecommunications industry.

   The existing literature has identified artificial intelligence as the solution for predicting customer churn. For example, machine learning (ML) and data analytics have facilitated customer churn prediction because ML algorithms could be used to collect large sets of customer data. Such customer data, which include the customer’s purchase history or trend, browsing behaviors, and interaction strategies, could be used to predict if a customer will leave the telecommunication services for another. However, existing studies show that these models can still fail or produce inaccurate results if applied to big data [20]. In contrast, deep learning (DL) models improve the prediction accuracy due to their ability to identify complex patterns in big data [21].For instance, both transformers and neural networks are capable of identifying complex trends that are missed by other approaches. These are useful insights for telecommunication companies that are aimed at retaining customers.

   Despite such progress in the use of emerging technologies and AI models for predicting customer churn, a gap exists in existing literature because current studies on customer churn prediction focus on traditional machine learning approaches involving decision trees, logistic regression, and random forests [19]. It has been noted that these ML models are limited in their capacity to capture complex relationships that naturally occur within customer behaviors [35]. More so, there are limited systematic literature reviews (SLR) that explain why customers churn from telecommunications company services. Furthermore, the few existing SLRs have not conceptualized the factors influencing customer churn in the telecommunications industry. On the contrary, the existing SLRs have concentrated on the challenges of retaining customers in the telecommunication industry [25].It was on

   this background that this study developed a conceptual framework of factors influencing customer churn in the telecommunications companies. For this purpose, this study answers the following questions:

   1. What key factors influence customer churn in the Telecommunications sector?
   2. .How is the customer churn currently predicted in the telecommunications industry?
2. METHODOLOGY

   An SLR was conducted following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta- Analyses (PRISMA). The following search string was used to search for peer-reviewed journal and conference articles published in the IEEE database, written in the English language, and published between the years 2023 and 2025. The initial search returned 754 articles after removing 10 duplicate articles. The screening process removed 726 articles to remain with only 28 relevant articles for inclusion in this SLR.

3. FINDINGS AND DISCUSSION

   Before you begin to format your paper, first write and save

   1. Factors Influencing Customer Churn

      identified include usage behavior (e.g., call/SMS/data usage, dropped calls), contract/service features (tenure, service type, tariffs), billing and payment patterns (late payments, etc.), and in some cases social network influences. Interestingly, studies focusing on network QoS [7] indicate that technical service quality metrics can also predict churn customers are more likely to leave if they experience poor network performance, a relevant insight for African operators dealing with infrastructure challenges. The table depicts the customer churn prediction factors and their categories.

      TABLE I. FACTORS INFLUENCING CUSTOMER CHURN

      Category	Factor	Source
      Customer Demographics	Age, gender, location, income level	[4] [1]
      Product/ servise use patterns	Call duration, data usage, SMS frequency	[5] [9], [35]
      Billing and Payment Be- havior	Late payments, arrears, contract type	[7]
      Network Quality (QoS	Call drop rate, latency, signal strength	[7] [2]
      Customer Int eraction Data	Complaint frequency, support calls, app usage	[16]
      Social Network Features	Number of churners in a users net-work	[9]

      It is evident from the findings that some factor categories are more pronounced in the literature. For example, customer demographics, product/service usage patterns, network quality, and customer interaction data. These emergent factor categories are depicted in Figure 1.

      FIGURE1 categories of prediction factors

      Figure 1 shows that customer churn is mostly influenced by customer demographics, product/service usage patterns, network quality, and customer interaction data, all at 20%, and least influenced by billing and payment behavior as well as social network features, which each represent 10% of the total categories.

      The additional moderating factors include those depicted in Table II.

      TABLE II. MODERATING FACTORS FOR CUSTOMER CHURN

      Moderator	Influence	Source
      Data Quality & Avail- ability	Limits the performance and generalizability of models	[2] [3]
      Regulatory & Ethical Constraints	GDPR compliance, privacy- preserving ML	[8] [11], [27]
      Business Integration & Profit Optimization	Links technical accuracy to financial outcomes	[11] [12]
      Regional Infrastructure	Affects data types and churn behavior (Africa vs EU)	[7]; [1]

      Popular among the moderating factors are data quality and availability, followed by regulatory and ethical constraints as well as business integration and profit optimization, while regulatory and ethical constraints are the unpopular factor. These observations are illustrated in Figure II.

      FIGURE II. categories of prediction factors

      Figure II demonstrates that data quality and availability (37%) have a central role in mediating customer churn, while regulatory and ethical constraints are the least influential (13%).

      The identified factors were then used to develop a conceptual framework of factors useful for predicting customer churn in the telecommunications industry. Figure III provides an overview of both the independent factors and the moderating factors that influence customer churn in the telecommunication industry.

      FIGURE III : CONCEPTUAL FRAMEWORK OF CUSTOMER CHURN FACTORS

   2. AI Strategies for Predicting Customer Churn

   Table III presents the different types of models that were referenced in the articles included in the SLR. The models included machine learning, ensemble, deep learning, explainable AI, support vector machine, transformer-based, natural language processing, random forest, decision tree, convolutional neural network, and artificial neural network.

   TABLE III. FACTORS INFUENCING CUSTOMER CHURN

   AI Models	Source
   Machine Learning	[6]; [3] [9]; [12]; [14], [15][2]; [1], [31], [32]
   Ensemble models	[6] [3] [9] [12]; [14] [15]; [2] [1], [24]
   Deep Learning	[3] [23]
   Explainable AI	[10]; [1], [33]
   Support Vector Machine	[6], [13], [31], [32]
   Transformer based	[16]
   Natural Language processing	[16]
   Random Forest	[1], [31], [32]
   Decision Tree	[1], [31]
   Convolutional Neural Network	[5] [2]
   Artificial neural network	[5]

   The most cited model for predicting customer churn was the Machine Learning model as illustrated in Figure IV.

   FIGURE I V.: AI MODELS FOR PREDICTING CUSTOMER CHURN

   The most popular model is machine learning (27%), followed by ensemble models (22%) and support vector machines (11%). Decision tree, natural language processing, random forest, transformer-based, and deep learning were not referenced much in the articles included in this study.

4. CONCLUSION

   In this study, an SLR was conducted to identify the factors influencing customer churn in the telecommunications industry. The study findings indicated that customer churn is a result of customer personal traits and the type of product/service. And network issues. The study further investigated the AI technologies used for predicting customer churn, and the most prominent were the machine learning models, ensemble models, support vector machines, explainable AI, random forests, etc. The study concluded by presenting a conceptual framework of factors influencing customer churn in the telecommunications industry. These findings are valuable to telecommunications companies that are at risk of losing customers and are eager to retain them. The study had limitations in that only two databases were searched.

   Similar future work could consider more databases for more customer churn factors to be identified.

   .REFERENCES

   1. Chang, V., Hall, K., Xu, Q. A., Amao, F. O., Ganatra, M. A. & Benson,
5. (2024). Prediction of Customer Churn Behavior in the Telecommunication Industry Using Machine Learning Models. Algorithms, 17(6), 231.

1. Fujo, S. W., Subramanian, S. & Khder, M. A. (2022). Customer churn prediction in telecommunication industry using deep learning. Information Sciences Letters, 11(1), 24.
2. Imani, M., Joudaki, M., Beikmohammadi, A. & Arabnia, H. R. (2025). Customer Churn Prediction: A System-atic Review of Recent Advances, Trends, and Challenges in Machine Learning and Deep Learning. Preprints, 202503.1969.v1.
3. Oladipo, I. D., Awotunde, J. B., AbdulRaheem, M., Taofeek-Ibrahim, F. A., Obaje, O. & Ndunagu, J. N. (2023). Customer Churn Prediction in Telecommunications Using Ensemble Technique. University of Ibadan Journal of Science and Logics in ICT Research, 9(1), 82-95.
4. Saha, C., Haque, M. M., Alam, M. G. R. & Talukder, A. (2024). ChurnNet: Deep Learning Enhanced Customer Churn Prediction in Telecommunication Industry. Proceedings of 2023 International Conference on AI in Engi-neering, IEEE (In press).
5. Sikri, A., Jameel, R., Idrees, S. M. & Kaur, H. (2024). Enhancing customer retention in telecom industry with machine learning driven churn prediction. Scientific Reports, 14, 13097.
6. Summaila, A. & Stephen, S. O. (2023). Performance Analysis of QoS Parameters and Churn Prediction Model Development for MNOS in Nigeria. Journal of Science Innovation & Technology Research, 2(2), 1- 12.
7. Wu, S., Yau, W. C., Ong, T. S. & Chong, S. C. (2021). Integrated churn prediction and customer segmentation framework for telco business. IEEE Access, 9, 62118-62136.
8. Andrews, J., Smith, R. & Johnson, K. (2022). Social Network Features for Enhanced Churn Pediction in Telecommunications. Journal of Business Analytics, 15(3), 45-62.
9. Verbeke, L., Wang, H. & Chen, X. (2022). XAI-Churn TriBoost: An Explainable Ensemble Approach for Customer Churn Prediction. Expert Systems with Applications, 198, 116857.
10. Verbeke, W., Dejaeger, K. & Martens, D. (2022). Profit-Driven Customer Churn Prediction Using Maximum Profit Performance Measures. European Journal of Operational Research, 218(2), 485-495.
11. Stripling, L., vanden Broucke, S. & Baesens, B. (2022). Profit-Based Model Evaluation for Customer Churn Prediction. Decision Support Systems, 153, 113668.
12. Khoh, W. H. (2023). Customer Churn Prediction through Attribute Selection Analysis and Support Vector Machine. Journal of Telecommunications and the Digital Economy, 11(1), 45-62.
13. Nagle, T. T. & Holden, R. K. (2022). The Strategy and Tactics of Pricing: A Guide to Growing More Profitably (6th ed.). Routledge.
14. Wedel, M. & Kamakura, W. A. (2022). Market Segmentation: Conceptual and Methodological Foundations (3rd ed.). Springer.
15. Semary, N. A., Tharvat, A., & Nasr, M. (2023). Transformer-based Models for Customer Churn Prediction from Textual Feedback. Journal of Artificial Intelligence Research, 76, 123-145.
16. Louro, A.C., Pugirá, C.G. and Murari, R.S. (2024) A scoping review for churn prediction: step-by-step tutorial and reproducible R code, Int. J. Business Forecasting and Marketing Intelligence, Vol. 9, No. 2, pp.160 178.
17. Pundru C. Shaker Reddy, Yadala Sucharitha and Aelgani Vivekanand (2024), Recent Advances in Electrical & Electronic Engineering,

    Volume 17, Issue 5, Jun 2024, p. 456 465 https://doi.org/10.2174/2352096516666230717102625

18. Sam, G., Asuquo, P., & Stephen, B. (2024). Customer churn prediction using machine learning models. Journal of Engineering Research and Reports, 26(2), 181-193.
19. Baghla, S., & Gupta, G. (2023, December). Data preprocessing for development of customer churn prediction models in e-commerce. In AIP conference proceedings (Vol. 2916, No. 1, p. 130002). AIP Publishing LLC.
20. Aldalan, A. M., & Almaleh, A. (2023). Customer Churn Prediction Using Four Machine Learning Algorithms Integrating Feature Selection and Normalization in the Telecom Sector. International Journal of Electronics and Communication Engineering, 17(3), 76-83.
21. Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gøtzsche, P. C., Ioannidis, J. P., … & Moher, D. (2009). The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. Bmj, 339.
22. Alboukaey, N., Joukhadar, A., & Ghneim, N. (2020). Dynamic behavior based churn prediction in mobile telecom. Expert Systems with Applications, 162, 113779.
23. Imani, M. (2025). Customer Churn Prediction in Telecommunication Industry: A Literature Review.
24. Onoja, J. P. A Comprehensive Review of Customer Churn Prediction Models in Telecommunications.
25. Saleh, S., Saha, S. Customer retention and churn prediction in the telecommunication industry: a case study on a Danish university. SN Appl. Sci. 5, 173 (2023). https://doi.org/10.1007/s42452-023-05389-6
26. Adekunle, B. I., Chukwuma-Eke, E. C., Balogun, E. D., & Ogunsola, K.

    O. (2023). Improving customer retention through machine learning: A predictive approach to churn prevention and engagement strategies. International Journal of Scientific Research in Computer Science,

    Engineering and Information Technology, 9(4), 507-523

27. Abdelhady, M. G., & Mohamed, K. A. (2025). Leveraging artificial intelligence for predictive customer churn modeling in telecommunications: a framework for enhanced customer relationship management. Scientific Reports.
28. Jahan, I., & Sanam, T. F. (2024). A comprehensive framework for customer retention in E-commerce using machine learning based on churn prediction, customer segmentation, and recommendation: I. Jahan, TF Sanam. Electronic Commerce Research, 1-44.
29. Asfe, A. M., Rahman, M. R., & Hossain, M. S. (2025). MNeuralTab: Integrating meta-modeling and neural networks for customer churn prediction in e-commerce. Discover Applied Sciences, 7(6), 569.
30. Kolli, M., Varadharajan, N., Ajith, K., & Kumar, K. D. (2024, March). Customer churn prediction in subscription-based services. In International Conference on Recent Trends in Machine Learning, IOT, Smart Cities & Applications (pp. 247-257). Singapore: Springer Nature Singapore.
31. Raju, P., Swathi, S., Sree, V. K. S., Durga, V. L., Niharika, P., & Pujitha, U. (2024, April). Telecom Customer Churn Prediction Using Machine Learning. In International Conference on Cognitive Computing and Cyber Physical Systems (pp. 207-220). Cham: Springer Nature Switzerland.
32. Yuan, J., & Liu, H. (2025, May). Telecom Customer Churn Prediction with Explainable Machine Learning. In Proceedings of the 2025 2nd International Conference on Modeling, Natural Language Processing and Machine Learning (pp. 246-250).
33. Liu, X., Xia, G., Zhang, X. et al. Customer churn prediction model based on hybrid neural networks. Sci Rep 14, 30707 (2024). https://doi.org/10.1038/s41598-024-79603-9
34. Kimitei, S., Agiro, D., Ni, S. et al. Predictability & explainability of survival analysis in churn prediction. J Market Anal (2025). https://doi.org/10.1057/s41270-025-00450-2