This study analyzes the influence of product attributes and customer reviews on sales performance on the Tokopedia e-commerce platform. Data were collected via web scraping from Tokopedia's electronics category (laptops and smartphones), yielding 463 products as the final sample. Product attributes (price, store status, brand clarity, and product description) and customer review indicators (review volume, average star rating, and sentiment score) were used as independent variables, while sales performance (sold count) served as the dependent variable. Sentiment analysis was conducted using a lexicon-based text mining approach with an Indonesian sentiment lexicon. Data analysis was performed using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS 4. Results indicate that product attributes have a positive and significant effect on sales performance (β = 0.141, t = 5.706, p < 0.05), and customer reviews have a highly significant effect (β = 0.823, t = 46.768, p < 0.05). Together, both variables explain 76.9% of the variance in sales performance (R² = 0.769). Customer reviews, particularly review volume, are the dominant determinant, while store status is the most influential product attribute indicator.
Copyrights © 2026