소비자간 정보 전달과 구전효과를 반영한 구조적 선택 모형 및 실증 적용

Abstract

Recently, the word-of-mouth effect that information on a specific product or service is transferred from some consumers to the others, is getting more and more important. Thesedays,in general, firms put a huge amount of investment and cost for developing new products or services. Therefore, if they fail in creating a positive word-of-mouth effect and cannot reach the critical mass point, it can lead to a huge loss not only to the firms associated with those products or services but also to the society. As such, many firms pay a lot of attention and put a lot of effort and resources to understand how the word-of-mouth effect among consumers takes place, and how to create the positive word-of-mouth effect for their products or services effectively, Also, from the academic perspective, these are considered to be very interesting research issues. A remarkable point is that in this word-of-mouth effect, normally a learning process is involved in that one consumer gets information on the quality of a specific product or service from his or her past experience, which means that he or she learns about the product or service (self-learning), and transfers that information or feeling to other consumer through the word-of-mouth, and this other consumer also learns about the product or service based on both his or her experience (self-learning) and other consumers' experience (learning from others). Therefore, to analyze how the mixture of word-of-mouth effect and the learning process affects the final choices of consumers, or to identify the determinant factors for them, is very significant in terms of consumer studies. However, even though some previous researches attempted to study this issue empiricially, most of them only modeled or analyzed this in a reduced-form manner. There are few studies that model the process related to the word-ofmouth effect and learning process more rigorously or theoretically, and attempt to analyze the consumer behavior quantitatively based on this more theoretically-proper model. This study is mainly aimed at proposing a structural choice model that can reflect the wordof-mouth effect in which information obtained through individual experiences of close consumers in specific groups is transferred and shared, and the learning process based on this information. Additionally, this study is aimed at examining whether the proposed model is valid by applying the model empirically. The proposed model is applied to the 4-year quarterly data containing 503 spouses' 1843choices of mobile-phone service providers in Korea. From the estimation results, compared to the other benchmark models (reduced-form mixed logit model, model without any learning process, model without learning from spouses), the proposed model is found to show better fitting statistics in terms of loglikelihood, AIC(Akaike Information Criterion), and BIC (Bayesian Information Criterion).