In the era of big data, information is one of one of the most valuable resources for enterprises. Hundreds of millions of enormous transactions have created rich data resources for the e-commerce platform. With the help of advanced information technology such as big data and cloud computing, platforms are able to extract valuable product information from massive commodity transaction records, and are closer to the demand side than ordinary manufacturers, having a natural information advantage. At the same time, by providing enabling services, the platform considers selling information to cooperative manufacturers to realize the value-added profit of data. However, not all platforms are willing to share product information with manufacturers for free, as information sharing may increase manufacturers′ profits while reducing their own profits. In addition, manufacturers can participate as sellers on e-commerce platforms and engage directly with the end market across the platform without relying exclusively on the platform to share information. In addition to demand information sharing decisions, the platforms also need to decide the optimal operation mode due to changes in the market environment. Therefore, it is a meaningful research question in the context of the e-commerce supply chain to develop proper information sharing strategies and operation modes for e-commerce platforms in a timely manner under different conditions.First, this study reviews and compares the relevant literature, including both the operation mode of e-commerce platforms and information sharing in supply chains. The results show that most of the previous articles studying supply chains via an analytical model approach focus on traditional supply chains with the default strength of manufacturers over retailers, and few scholars have endogenized the service-level of e-commerce platforms for decision making. Previous studies on information sharing have mainly focused on traditional supply chains or dual-channel supply chains, usually ignoring information costs and taking retailers′ demand information advantage as the existing premise. We propose a “pyramid” logical structure, in which three levels of management decisions are studied from the bottom up, starting with solving the Stackelberg game model to obtain operational decisions at the executive level, then analysing the information sharing decisions at the tactical level, and finally rising to the operation mode selection decisions at the strategic level.Second, based on the e-commerce supply chain, considering both platform service-level decisions and information collection and acquisition costs, we investigate the optimal operational decisions and demand information sharing decisions through Stackelberg game model where the e-commerce platform acts as the dominant player and the manufacturer as the follower under two modes———reseller mode and marketplace mode. The results show that each equilibrium decision is mainly influenced by the platform′s service performance. Information sharing decisions depend on the accuracy of forecast information and costs. Both the manufacturer and platform make trade-offs between the value and cost of information sharing. It is easier to achieve information collaboration under the reseller mode. Specifically, in the reseller mode, as the system becomes less efficient, the platform is unwilling to maintain a higher service level, and manufacturer and platform will reduce prices to stabilize sales; however, the final product′s demand will still decline and will shrink to only a quarter of the potential demand in the case of very poor efficiency. In the marketplace mode, as the platform′s service performance decreases, the optimal product price and service-level also decrease, and demand shrinks to one-half of the potential demand.Finally, this study compares the optimal product selling price, service-level and expected profit under different modes. The results show that the optimal product selling price and service-level under the reseller mode is consistently higher than those under the marketplace mode, regardless of whether information is shared or not. When service cost performance is at a higher level, the manufacturer′s expected profit is higher under the reseller mode, while the platform always prefers the reseller mode. This counterintuitive finding can be attributed to both the e-commerce platform and the manufacturer benefit from the larger market demand and higher service price only when the platform′s service cost performance is high, because the positive effect of service level increase on demand is greater than the negative effect of price increase on demand. Only under these conditions can both parties choose to collaborate under the reseller mode. This conclusion aligns with actual operational practices observed in self-operated businesses such as JD.com, Tmall, and Meituan.In summary, with the rapid advancement of e-commerce, e-commerce platforms with the advantage of demand information gradually dominate supply chain, making their choices regarding information sharing strategies and operation modes critical determinants. Our findings offer valuable managerial insights and practical guidance, facilitating these platforms to fully exploit their informational advantages and dominant vitality while enabling them to make scientifically grounded management decisions aimed at maximizing enterprise profitability.