In today’s era of rapid digital development, the competition among retail stores has become increasingly fierce, and sales forecasting plays a pivotal role in this competitive landscape. Large retail stores, with their extensive range of products and diverse geographical locations of branches, face significant challenges in accurate sales prediction. Traditional prediction methods often fall short due to their limitations, such as insufficient data processing capabilities and lack of adaptability to dynamic market conditions, leaving many retailers unable to make timely and precise forecasts. To address this issue, this study conducted a comprehensive evaluation by implementing and comparing multiple predictive models, including traditional statistical methods and advanced machine learning techniques. The research findings demonstrate the superior accuracy and effectiveness of deep learning models in sales forecasting. The final model developed in this study, which incorporates deep learning algorithms, can significantly enhance retailers’ ability to predict future sales with greater precision, thereby enabling more informed decision-making and strategic planning. This improved forecasting capability can help retailers optimize inventory management, reduce operational costs, and improve customer satisfaction.