Soft Computing Modeling of Dealer Loyalty in Turkish Insurance Sector

This study presents the application of soft computing techniques namely as Stepwise Regression (SR), Neuro-Fuzzy (NF) and Neural Networks (NN) for modeling of dealer loyalty in Turkish insurance sector. The proposed soft computing models are based on survey results conducted in insurance sector of Turkey. The accuracies of the proposed soft computing models are quite satisfactory as compared to actual results. The results show that the dominating factors affecting dealer loyalty are found to be dealer satisfaction, dealer trust, dealer perceived value, dealer perception of customer perceived value and relationship time. The analyses of effects of research variables on dealer loyalty are an important contribution to understanding relationships between dealer and manufacturer. Finally, theoretical and managerial implications of the study findings are discussed.