온라인 구매가 활발해지면서 평점과 후기는 소비자의 구매 결정에 영향을 미치는 주요 정보로 이용되고 있다. 본 연구에서는 탐색재와 경험재의 차이로 주로 언급되는 경험재 평가에서의 주관성과 개인의 취향이 반영되는 점을 경험적으로 검증해보고자 하였다. 탐색재의 경우 객관적인 품질 평가가 가능하여 평점과 후기의 영향이 개인의 취향이나 선호와 상관없이 비교적 일관되게 나타나지만, 경험재는 제품 평가가 주관적이므로 개인의 취향 또는 선호에 따라 평점과 후기의 영향이 달라질 수 있다. 후기 내용이 경험재를 평가하는 데 도움이 되는 사용 경험과 개인의 취향에 대한 것일 때 후기가 탐색재보다 경험재에 더 영향을 미치는지 살펴보고, 선호 차별화와 선호 유사성을 경험재 평가에 영향을 미칠 수 있는 선호 관련 변인으로 제시하여 선호 차별화 정도에 따라 경험재에 미치는 평점의 영향이 다르게 나타나는지, 후기 유형에 따라 후기가 구매 의도에 미치는 영향에 선호 유사성과 선호 차별화의 조절 효과가 다르게 나타나는지 알아보고자 한다. 결과를 보면, 후기의 두 유형이 모두 경험재 평가에 도움이 되는 내용이기에 탐색재보다 경험재의 구매 의도에 미치는 후기의 영향이 더 큰 것으로 나타났고, 대중의 지혜를 나타내는 평점의 경우 선호 차별성이 낮을 때만 경험재 구매 의도에 영향을 미쳤다. 두 후기 유형 중 개인의 취향에 대한 후기의 경우 리뷰어와의 선호 유사성과 제품군에 대한 선호 차별성이 후기가 구매 의도에 미치는 영향을 조절함을 볼 수 있었다. 즉 선호가 유사할수록 후기가 구매 의도에 미치는 영향이 커지지만 선호 차별성이 커지면 선호 유사성의 조절 효과가 줄어들었다. 이러한 조절된 조절 효과는 경험재에 대해서만 나타났다. 마지막으로 결과의 시사점을 논의하였다.
As online purchases become more prevalent, ratings and reviews are used as key information that influences consumers’ purchase decisions. This study tried to empirically verify that subjectivity and personal taste are reflected in the evaluation of experience goods, which have been mainly mentioned as the difference between search goods and experience goods. In the case of search goods, objective quality evaluation is possible, so the influence of ratings and reviews appears relatively consistent regardless of personal taste or preference. In case of experience goods, however, the influence of ratings and reviews may vary depending on personal taste or preference because quality evaluation is subjective. This study aims to examine whether the review affects the experience goods more than the search goods, whether the effect of ratings on experience goods differs depending on the degree of preference differentiation, and whether the moderating effects of preference similarity and preference differentiation on purchase intention differ depending on the type of reviews when the review content is about the usage experience and personal taste that help evaluate the experience goods and preference differentiation and preference similarity are presented as preference related variables.
The experiment was conducted in the mixed design of product type (search vs. experience goods) × review type (usage experience vs. personal preference) × review direction (positive vs. negative) × purchase intention (before review vs. after review). The first three variables are between-subjects variables and the last variable is a within-subjects variable. Based on the way products are presented in most online shopping malls in Korea, ratings and product information were first presented and purchase intention (purchasing intention 1) was measured. Then, reviews were presented and purchase intention (purchase intention 2) was measured again (refer to the appendix). There were conditions with high ratings (4.5 stars, 2,845 reviews) and conditions with low ratings (2 stars, 2,845 reviews). The rating was not taken as a separate variable because the rating and the review direction were reversed.
To find out proper product types and review contents, three pretests were conducted online. Depending on the specific contents of the reviews, the helpfulness of the reviews and the expertise of the reviewers were perceived differently, so these variables were controlled as covariates. The main experiment was conducted online for the Embrain panel (survey period: March 7~11, 2022). Four hundred forty eight men and women in their 20s and 30s who are familiar with online shopping participated, and 56 people were assigned to each condition randomly. Data from 434 participants were analyzed, excluding the 14 participants who had the product presented as stimuli. There were 215 males (49.5%) and 219 females (50.5%), and the mean age was 30.02 years (SD=5.27).
First, a 2 × 2 × 2 × 2 repeated measurement ANOVA was conducted to test Hypothesis 1 examining the effect of reviews on changes in purchase intention. Looking at the product type × purchase intention interaction, the difference between purchase intention before and after was little in the case of search goods (before M=3.44, SE=.10 vs. after M=3.42, SE=.09). On the contrary, in the case of experience goods, purchase intention decreased significantly after presenting a review (before M=3.90, SE=.10 vs. After M=3.61, SE=.09, p<.001), supporting Hypothesis 1. To verify Hypothesis 2, Hayes(2013)’s PROCESS analysis was performed (Model 3). The independent variable was rating (low: 0, high: 1), the dependent variable was purchase intention 1, and the moderating variables were product type and preference differentiation. The three-way interaction was significant, supporting Hypothesis 2 stating that the effect of ratings affected the purchase intention of experience goods only when preference differentiation was low. To test Hypothesis 3, Hayes(2013)’s PROCESS analyses were performed separately for four different conditions of product type and review type. The independent variable was review direction (positive: 0, negative: 1), the dependent variable was changes in purchase intention (purchase intention 2 - purchase intention 1) and the moderating variables were preference similarity and preference differentiation. The three-way interaction was significant, supporting Hypothesis 3. That is, the more similar the preference, the greater the effect of reviews on purchase intention, but the greater the preference differentiation, the less the moderated effect of preference similarity. This moderated moderating effect was shown only for the experience product and the review type of personal preference.
The most important factor influencing purchase decision is the specific content of the review. This study focused on the specific content of the review, which includes the usage experience and personal preference that can help the evaluation of experience goods. The results suggest that it is necessary to look at the influence of preference-related variables together with specific content details because individual preferences or tastes affect quality evaluation or purchase decisions of experience goods. In the future, it seems that variables representing content characteristics should be proposed and the effect of such characteristics on the helpfulness of reviews should be studied more actively. In practice, if you find out these content characteristics, have customers who write reviews evaluate them quantitatively, and show them to customers, online purchases will be much easier. In particular, in the case of experience goods, past purchase information or review information that can determine the preference similarity with reviewers will be of great help to purchase decisions.