.K. Latha
Research Subjects
.K. Latha has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- .B. Kiruthiga
Assistant Professor, Department of Management Studies, SRM Valliammai Engineering College, Kattankulathur, Tamil Nadu, India
1 shared publication - .V. Suganya
Assistant Professor (SG), Faculty of Management-MBA, SRM Institute of Science and Technology, Vadapalani Campus, Chennai, Tamil Nadu, India
1 shared publication - Priya Sethuraman
Professor, Department of Management Studies, St. Joseph’s Institute of Technology, OMR, Chennai, Tamil Nadu, India
1 shared publication
Latest Publications
- Open Access
- 29 September 2026
- Published by National Institute of STEM Research in Archives for Technical Sciences
- No. 36,p. 1154-1167
VIRTUAL CONSUMER TWIN FRAMEWORK FOR PREDICTIVE MARKETING AND CUSTOMER RETENTION
Organizations that want to gain a competitive advantage have focused on customer retention. However, traditional approaches to customer relationship management do not account for changing customer behaviour in the digital realm. This work suggests the Virtual Consumer Twin Framework (VCTF) for customer retention and predictive marketing through delivering behavioural analytics, customer segmentation, predictive modeling, and the process of continuous updating of digital twins. It seeks to achieve higher accuracy in terms of retention predictions and to improve personalized means of marketing through being able to engage with customers online in real time. A quantitative research design was utilized for examining transaction, demographic, and online engagement data from 1,320 individuals over the period of 36 months. Thus, Structural Equation Modeling (SEM) was adopted to study the connections between digital engagement, quality of personalization, customer satisfaction, customer trust, efficiency of predictive marketing activities, and retention, whereas predictive validation was based on past behaviour and binary classification indicators. The model suggested in the study was successful in predicting with 94.1% accuracy, AUC being 0.952, precision at 92.8%, recall rate being 91.6%, the score of F1 was 92.2%, and RMSE is 0.182, and MAPE equals 5.12%. After SEM analysis has been conducted it was revealed that personalization quality has a positive influence on consumer satisfaction (β=0.81, p<0.001), satisfaction has a positive influence on client trust (β=0.76, p<0.001), and trust influences customer retention (β=0.78, p<0.001), the efficiency of marketing efforts provides customer retention (β=0.84, p<0.001), and the level of digital engagement influences the efficiency of marketing efforts (β=0.79, p<0.001). The results showed that the measurement model demonstrated good reliability and validity, with CFI being 0.957, TLI being 0.949, RMSEA being 0.039, SRMR = 0.036, and Cronbach’s α being 0.84 to 0.92. The data confirm that Virtual Consumer Twin (VCT) technologies provide a solid framework for analysing customer behaviour through data-based personalization that enables companies to make better marketing decisions, strengthen ties with customers, and enhance long- term customer retention.
- 29 September 2026
- Published by National Institute of STEM Research in Archives for Technical Sciences
- No. 36,p. 1154-1167
Open AccessVIRTUAL CONSUMER TWIN FRAMEWORK FOR PREDICTIVE MARKETING AND CUSTOMER RETENTION
Organizations that want to gain a competitive advantage have focused on customer retention. However, traditional approaches to customer relationship management do not account for changing customer behaviour in the digital realm. This work suggests the Virtual Consumer Twin Framework (VCTF) for customer retention and predictive marketing through delivering behavioural analytics, customer segmentation, predictive modeling, and the process of continuous updating of digital twins. It seeks to achieve higher accuracy in terms of retention predictions and to improve personalized means of marketing through being able to engage with customers online in real time. A quantitative research design was utilized for examining transaction, demographic, and online engagement data from 1,320 individuals over the period of 36 months. Thus, Structural Equation Modeling (SEM) was adopted to study the connections between digital engagement, quality of personalization, customer satisfaction, customer trust, efficiency of predictive marketing activities, and retention, whereas predictive validation was based on past behaviour and binary classification indicators. The model suggested in the study was successful in predicting with 94.1% accuracy, AUC being 0.952, precision at 92.8%, recall rate being 91.6%, the score of F1 was 92.2%, and RMSE is 0.182, and MAPE equals 5.12%. After SEM analysis has been conducted it was revealed that personalization quality has a positive influence on consumer satisfaction (β=0.81, p<0.001), satisfaction has a positive influence on client trust (β=0.76, p<0.001), and trust influences customer retention (β=0.78, p<0.001), the efficiency of marketing efforts provides customer retention (β=0.84, p<0.001), and the level of digital engagement influences the efficiency of marketing efforts (β=0.79, p<0.001). The results showed that the measurement model demonstrated good reliability and validity, with CFI being 0.957, TLI being 0.949, RMSEA being 0.039, SRMR = 0.036, and Cronbach’s α being 0.84 to 0.92. The data confirm that Virtual Consumer Twin (VCT) technologies provide a solid framework for analysing customer behaviour through data-based personalization that enables companies to make better marketing decisions, strengthen ties with customers, and enhance long- term customer retention.