Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM68468
Volume:
10 |
10 |
Oct 2026
Pages:
1-20
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
School of Management
Suresh Gyan Vihar University, Jaipur
School of Management
Suresh Gyan Vihar University, Jaipur
School of Management
Suresh Gyan Vihar University, Jaipur
School of Management
Suresh Gyan Vihar University, Jaipur
School of Management
Suresh Gyan Vihar University, Jaipur
School of Management
Suresh Gyan Vihar University, Jaipur
Purpose. Banks in the United Arab Emirates (UAE) are investing heavily in AI-based chatbots, yet managers still lack evidence on which customer behaviours a chatbot can realistically change and how much those behaviours matter for retention. This study isolates the behavioural signals that precede attrition and quantifies the share of attrition risk that a conversational agent can address. Design. A secondary, quantitative design analyses two independent open retail-banking datasets: a primary sample of 10,127 credit-card customers (1,627 attrited) and a replication sample of 10,000 bank customers (2,037 exited). Five chatbot-addressable engagement levers were operationalised: inactivity, service-contact intensity, transaction frequency, engagement momentum and relationship breadth. The analysis combines Wilson confidence intervals, chi-square tests with Cramér's V, Welch t-tests with Cohen's d, nested binary logistic regression, a composite Digital Engagement Index (DEI), bootstrap what-if scenarios and a two-axis retention priority map. Findings. Demographics barely separated leavers from stayers (Cramér's V ≤ 0.037), whereas the engagement levers did (V = 0.17–0.21; |d| up to 1.09). Adding the five levers to a controls-only model raised McFadden's pseudo-R² from 0.070 to 0.405 and the AUC from 0.693 to 0.899 (LR χ²(5) = 2,987, p < .001). Each standard-deviation rise in service contacts raised the odds of attrition by 69% and each rise in inactivity by 61%. Attrition fell from 68.1% in the least-engaged DEI decile to 1.1% in the most-engaged decile. The pattern replicated in the second bank (active members churned at 14.3% versus 26.9%; AUC 0.752 → 0.834). A combined chatbot programme implied 636 avoided attritions per 10,000 customers (95% CI 601–672), and 29% of customers in the two engagement-risk segments accounted for 80% of all attriters. Originality and value. The study contributes a Friction–Disengagement Matrix, a transparent DEI and a Chatbot Retention Priority Map that tell a bank where a chatbot should act alone, where it should work with a relationship manager, and where it should step aside. The evidence is translated into a staged capability roadmap for UAE retail banks.
KEYWORDS
This study asked which customer behaviours a banking chatbot can influence, how much they matter for retention, and how a bank should act on them. In two independent retail banks, behaviour explained far more attrition than demographics. Service friction and inactivity raised the odds of leaving by 69% and 61% per standard deviation and amplified each other; usage frequency, momentum and moderate relationship breadth protected the relationship. The five levers raised explained variance almost six-fold in the primary bank and more than doubled it in the replication bank. A transparent index of the same levers placed 77% of attriters in the least-engaged 30% of customers, and a combined chatbot programme implied about 40% fewer attritions under stated assumptions. All five sub-hypotheses of the programme's retention hypothesis were supported, with breadth showing a ceiling.
[1] Huang, Ming-Hui, and Roland T. Rust. "Artificial Intelligence in Service." Journal of Service Research, vol. 21, no. 2, 2018, pp. 155–172. [2] Verhoef, Peter C., et al. "Digital Transformation: A Multidisciplinary Reflection and Research Agenda." Journal of Business Research, vol. 122, 2021, pp. 889–901. [3] Adamopoulou, Eleni, and Lefteris Moussiades. "Chatbots: History, Technology, and Applications." Machine Learning with Applications, vol. 2, 2020, article 100006. [4] Rahman, Mahfuzur, et al. "Adoption of Artificial Intelligence in Banking Services: An Empirical Analysis." International Journal of Emerging Markets, vol. 18, no. 10, 2023, pp. 4270–4300. [5] Qureshi, Omer Aftab, et al. "Chatbots: Can They Satisfy Customers in the Banking Sector?" BUiD Doctoral Research Conference 2023, edited by Khalid Al Marri et al., Lecture Notes in Civil Engineering, vol. 473, Springer, 2024, pp. 360–376. [6] Jenneboer, Liss, et al. "The Impact of Chatbots on Customer Loyalty: A Systematic Literature Review." Journal of Theoretical and Applied Electronic Commerce Research, vol. 17, no. 1, 2022, pp. 212–229. [7] Ascarza, Eva. "Retention Futility: Targeting High-Risk Customers Might Be Ineffective." Journal of Marketing Research, vol. 55, no. 1, 2018, pp. 80–98. [8] Hollebeek, Linda D., et al. "Rise of the Machines? Customer Engagement in Automated Service Interactions." Journal of Service Research, vol. 24, no. 1, 2021, pp. 3–8. [9] Kumar, V., et al. "Understanding the Role of Artificial Intelligence in Personalized Engagement Marketing." California Management Review, vol. 61, no. 4, 2019, pp. 135–155. [10] Eren, Berrin Arzu. "Determinants of Customer Satisfaction in Chatbot Use: Evidence from a Banking Application in Turkey." International Journal of Bank Marketing, vol. 39, no. 2, 2021, pp. 294–311.
Publication Date
10/29/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
8
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...