Abstract:
Objective With the rising demand for high heel shoes and the prevalence of online shopping as the mainstream consumption method, there is a large amount of unexplored consumer preference and pain point information hidden in online reviews of high heel shoes. This study aims to analyze online high heel shoes reviews to identify the key factors that consumers consider in their purchasing decisions, providing data support for enterprises and merchants to enhance product competitiveness and customer satisfaction.
Method The study collected 38,199 valid reviews of the top 200 best-selling high heel shoes on Tmall. Text mining methods were comprehensively applied: the Term Frequency-Inverse Document Frequency algorithm was used to extract high-frequency keywords and create word clouds; the KH Coder was used to construct a keyword co-occurrence network; the SnowNLP library was used for sentiment classification to identify positive and negative reviews as well as sentiment transition types; and the LDA topic model was used to cluster positive and negative reviews separately, generating five themes for each and obtaining their corresponding keyword lists.
Result Word frequency and co-occurrence analysis indicated that "good-looking" and "comfortable" were the two most concerned dimensions by consumers, which together formed the core basis of satisfaction. Sentiment analysis showed that approximately 84.34% of the comments were positive, but 15.66% of consumers still expressed dissatisfaction. Follow-up comments were found to be more informative than initial comments. Among them, "positive-to-negative" comments, although accounting for a small proportion (1.02%), revealed hidden problems such as foot chafing and glue detachment during product use. In contrast, "negative-to-positive" comments highlighted the remedial value of high-quality customer service. LDA topic clustering further classified positive reviews into five themes: external quality, wearing experience, product service, product quality, and product discounts. Negative reviews focused on external quality, wearing discomfort, service dissatisfaction, quality defects, and size inaccuracy.
Conclusion Consumers' demand for high heel shoes shows a dual characteristic of "emphasizing both appearance and comfort". Current market products have gained more recognition in terms of appearance, but there is still much room for improvement in comfort, workmanship quality, and size fitting. Follow-up reviews and the sentiment transitions within them are important sources for discovering potential product issues. Manufacturers should optimize shoe lasts based on ergonomics and real foot shape data, enhance heel structure stability, and improve the bonding process of uppers and soles. Online merchants should introduce intelligent size recommendation tools and establish active tracking and compensation mechanisms for negative customer reviews to enhance the post-purchase experience and brand loyalty. Future research can be extended to the analysis of high heel shoes reviews across multiple platforms, multi-price range, or specific functional categories.