随着电子商务的高速发展, 如何通过最优商品定价决策来帮助企业获得最优利润已成为网络零售商运营管理的重要研究方向. 随着促销活动的频次和种类的不断增加, 商品存在促销关联效应, 不同的促销策略会对商品间的销量产生影响. 首先, 应用 Granger-LASSO(least absolute shrinkage and selection operator)检验分析存在促销关联效应的商品; 然后, 使用 ADL(advanced distributed learning)-LASSO 多阶段回归建立商品需求函数模型; 最后, 结合利润函数, 构建出一种考虑促销关联效应的动态定价模型. 模拟结果表明, 该模型将 U 品牌商的线上销售总利润提升了 13.46%, 是解决网络零售商动态定价决策的有效方案.

In addition to the rapid development of e-commerce business, helping enterprises obtain optimal profits through dynamic pricing decisions has become a critical research direction of operations management for online retailers. With the increase in the frequency and variety of promotional activities, there are increasingly strong promotional correlation effects amid products. Therefore, different promotional strategies impact sales at the inter- and intra-category level. This study first applies the Granger-LASSO (least absolute shrinkage and selection operator) test to identify promotional correlation effects. The ADL (advanced distributed learning)-LASSO multi-stage regression is then used to establish a demand forecast model. Finally, the demand model is integrated into the profit function to establish a dynamic pricing model that considers promotional correlation effects. The experimental simulation results indicate that the proposed model can improve the total online profit of Brand U by 13.46%, which is proven to be an effective solution to dynamic pricing decisions for online retailers.