Predicting complaint voicing or exit amidst Indian consumers: a CHAID analysis

Abstract

PurposeThe current study aims to predict consumer complaint status (complainers or non-complainers) based on socio-demographic and psychographic factors and further to discern the differences in behavior disposition of consumer groups concerning determinants of consumer's tendency to exit (TE).Design/methodology/approachThe research used survey-based data of 600 Indian consumers of three service sectors (hotel and hospitality, automobile service centers and organized retail stores). Chi-square automatic interaction detector (CHAID) decision tree analysis was used to profile consumers.FindingsThe results indicated that occupation; income; education; industry and attitude toward complaining were significant factors in profiling consumers as complainers or non-complainers. Further, determinants of TE (discouraging subjective norms, perceived likelihood of successful complaint, lower perceived switching cost, poor employee response, negative past experience and ease of complaint process) vary significantly across the groups of complainers and non-complainers.Research limitations/implicationsThe research questions in this study were tested with three service sectors consumers in India, so due care should be exercised in generalizing these findings to other sectors and countries. Study replication across other service sectors and countries is recommended to improve the generalizability of these findings with wider socio-demographic samples.Practical implicationsFirms striving for consumer retention and aim to extend their consumer life cycle can greatly benefit from the results of this study to understand the customer complaint behavior (CCB) specific to non-complaining (exit) behavior. The future researcher may benefit from replicating and extending the model in different industries for further contribution to the CCB literature.Originality/valueTo the best of the author's knowledge, there is no evidence of consumer segmentation based on their complaining behavior or socio-demographic and psychographic factors by employing CHAID decision tree analysis. In addition to illustrating the use of data mining techniques such as CHAID in the field of CCB, it also contributes to the extant literature by researching in a non-Western setting like India.

Main motivations and barriers to pro-environmental behaviour: a study from the employee’s perspective

“…This analysis is expected to provide heterogeneous segments that differ significantly in terms of independent and dependent variables, as well as other variables. CHAID methodology has been extensively validated in the literature in different fields of study (Pessach et al , 2020; Surucu-Balci et al , 2020; Patalas-Maliszewska et al , 2022; Kumar and Kaur, 2023) and more specifically, in previous studies analysing perceptions, attitudes and behaviour towards sustainability and climate change (Shinkevich et al , 2021; De Luca et al , 2022; Ferreira et al , 2022; Juma‐Michilena et al , 2023).…”

“…The CHAID methodology determines the predictor variables that explain to a greater extent the differences between segments for the dependent variable. These segments are generated through a descending iterative method that allows the division of the initial sample into homogeneous and selective subsets, permitting the visual representation of the relationship found between the data in the form of a decision tree (Kumar and Kaur, 2023). This technique is structurally based on regression and classification models for analysis of multiple effects that make it possible to explain, describe or predict a result (Gunduz and Al-Ajji, 2022).…”

Improved Chi Square Automatic Interaction Detection on Students Discontinuation to Secondary School

“…Categorical data are generally analyzed using the Chi-Square Test to determine the relationship between variables. CHAID, with the Chi-Square Tests help, will form a classification (Yang et al, 2023;Kumar and Kaur, 2023). According to Collins (2021), CHAID is a method for classifying categorical data in which the purpose of the procedure is to divide the data set into sub-groups based on the dependent variable.…”