This study aims to evaluate the model's performance in classifying and predicting the number of product sales based on several attributes. The K-Nearest Neighbors (KNN) algorithm was used for the classification task and showed good performance with an accuracy of 91.22%, recall of 92.96%, and precision of 90.18%. These results indicate that the model has a high generalization ability in recognizing sales patterns. For quantitative prediction of the number of sales, a linear regression model is used with independent variables such as regular price, selling price, rating, number of ratings, and number of favorites. The regression model yielded a coefficient of determination (R²) of 0.78, indicating that 78% of the variability of the sales amount can be explained by these variables. The coefficient analysis results show that the rating, number of ratings, and number of favorites have a positive influence on the sales amount, while the selling price and normal price have no significant effect. The Root Mean Squared Error (RMSE) value of 33.09 indicates a fairly low level of prediction error. These findings indicate that the model used is effective in helping analyze and forecast product sales.