With the rapid development of modern information society, all types of advertisements emerge in an endless flow, where ambient media that emphasizes the multi-sensory experience of users rise as the times require. In this paper, based on the price stepping algorithm, a new ambient media advertising model based on O2O that integrates offline large-screen resources is proposed to further promote the development of ambient media in modern society. By introducing the price stepping (PPA) algorithm for sales and management analysis and sensitivity evaluation of actual goods, we aim to dynamically solve the marketing and inventory decision problem of environmental media advertising products, and on this basis, we use the price stepping (PPA) algorithm to make sales forecasts for high-cost goods, provide decision support for inventory and theoretical balance, and achieve theoretical maximization of environmental media advertising. The simulation results show that the PPA algorithm provides a more accurate sales forecast for high-cost goods, which in turn provides a joint decision idea to achieve inventory optimization and profit maximization, and provides an optimization method for merchants to balance inventory costs and marketing costs to maximize profits.