We present a next-purchase recommendation system that combines advanced algorithms with explainable AI (XAI) to learn individual customer preferences from purchase histories and deliver personalized recommendations that enhance user engagement and inform marketing strategy. Our approach provides dual-layer, multistakeholder explanations: targeted communications that promote personalized marketing messages for customers and strategic insights for business stakeholders (e.g., marketing departments), reducing cognitive load and fostering trust. The system also addresses cold-start scenarios and leverages implicit feedback. Experiments on the MovieLens dataset demonstrate a balanced trade-off between accuracy, novelty, and explainability, potentially lowering users' decision-making effort.