Mass customization allows companies to offer personalized products efficiently, but demand forecasting remains challenging due to high variability in customer preferences. Unlike mass production, where historical sales data provides stable demand trends, mass customization requires modeling dynamic customer-product interactions. However, no publicly available dataset supports such analysis. This study generates and validates a synthetic dataset of 500,000 sales transactions in the electronics industry, focusing on mass-customized laptops. The dataset captures customer segmentation, customization preferences, and demand variability. To ensure its reliability, we applied feature distribution analysis, logical consistency checks, and demand pattern validation.