The Influence of Internal and External Factors on Readers' Evaluation in Word-of-Mouth Reviews of Lotion
Satoru Imoto (AY 2022)
In recent years, many cosmetic products have been sold in stores, and people buy cosmetics according to their purposes. In particular, basic cosmetics can cause skin problems if used with products that are not compatible with the skin, so people use word-of-mouth websites to select products that are compatible with their skin. However, the compatibility of basic cosmetics and skin varies from person to person, and it is difficult to determine the similarity between the skin of a viewer and a poster before purchasing a product. Although there have been previous studies on word-of-mouth recommendation systems that recommend products to viewers, we believed that the high degree of freedom in word-of-mouth expression limits the accuracy of the recommendation process. Therefore, in this study, we focused on word-of-mouth, which is the basis of the recommendation system, and investigated the effects of different word-of-mouth creation instructions on word-of-mouth viewers, with the goal of generating useful word-of-mouth for viewers.
The proposed instructions were created using three measures of similarity, credibility, and expertise that define the usefulness of a communication framework. For similarity and expertise, we created instructional content that could describe both external and internal factors, based on the purpose-related similarity defined by Simons et al. and the fact that both are "trust in the competence of the information source" factors. For credibility, we used four of the six axes created by Hamano et al. as axes for evaluating the credibility of word-of-mouth communication, omitting axes unrelated to judging whether or not the product fits the skin and axes that limit word-of-mouth expression. Lotion, which has the largest number and variety of products among basic cosmetics, was selected as the target of word-of-mouth communication. The experimental design was based on the proposed method and an existing method (at-cosmetics), and consisted of a step-by-step experiment in which participants were asked to "write word-of-mouth messages based on the instructions presented" and an experiment in which they were asked to "evaluate the written word-of-mouth messages". In the word-of-mouth experiment, 20 female students aged 20 or older who had never posted a word-of-mouth message were asked to write 170 to 230 words about a lotion they had been using for at least one month. In the word-of-mouth evaluation experiment, 28 women in their 20s were recruited through a crowdsourcing service and asked to evaluate the word-of-mouth messages obtained in the word-of-mouth description experiment.
The profile analysis and statistical analysis of the word-of-mouth evaluation scores obtained in the evaluation experiment showed that the word-of-mouth generated by the proposed method had a higher similarity score to viewers than the existing method, the difference was significant, and the effect size was also large. In terms of usefulness, the proposed method had a higher usefulness score for readers than the existing method, the difference was significant, and the effect size was moderate. No effect of differences in instructional content was observed for credibility, expertise, and purchase intention. This may be due to the ambiguity of the expression of credibility in the instructions and the expression of purchase intention in the evaluation items, and thus the effect of instructional content was not observed in some areas. These results suggest that the similarity and usefulness of generated word-of-mouth messages to viewers can be improved by including both external and internal factors in the generation of word-of-mouth messages. In the future, it is expected to be applied to other categories of basic cosmetics.
(Translated by DeepL)