An Interpretable Optimal Pricing Strategy for E-Commerce Platforms

Strategic and optimal pricing plays a significant role in ensuring profitability in the recent e-commerce world. Due to fierce competition and the dynamic nature of the market, pricing decision-making has become highly challenging. The manual methods of pricing are inefficient and prone to generating suboptimal solutions. However, the availability of large volumes of data generated by e-commerce platforms calls for a data-driven approach. With the improvement in statistics and machine learning methods, it is now possible to make informed decisions based on historical data that will ensure optimal pricing and increased profit. In this study, we propose an interpretable LSTM-XGBoost-based optimal pricing strategy that simultaneously forecasts the relevant variables for the projected time-window and optimizes the pricing decision based on correlations between potential price points and relevant dependent variables. The computational complexity of the proposed approach has been critically discussed and compared with the existing reinforcement learning-based method. Empirical results of the live testing phase indicate that the proposed model can increase the profit and lead to an upward trend in the profit curve.