;How to balance inventory and sales demand has always been a topic in the supply chain in target of low inventory and high profit. The chemical manufacturing industry is highly dependent on raw material imports in Taiwan due to the resource constraints of Taiwan′s congenital island. The severity of the epidemic impacted by COVID-19 outbreak in 2019. Various countries have successively published restriction policies such as city closures and border control that global traffic status has been interrupted. Manpower issues have disrupted the movement of supplies around the world which caused supply chain’s difficulty to adjust in time resulting in a bullwhip effect. In this unstable situation, a large number of logistics and inventory costs are relatively increased, and it also highlights the importance of supply and demand forecasting for enterprises, because enterprises′ raw material or finished product inventory is based on order demand. Therefore, this research will analyze the forecasting methods of the internal order demand of enterprises, compare the forecasting accuracy of the three forecasting methods of moving average method, exponential smoothing method, and neural network method, and find out the tool optimization forecasting method which is suitable for enterprise order forecasting. Strategic analysis to forecasting method and company′s production type with more effective model can reduce inventory costs and the impact of the bullwhip effect, and improve supply chain performance.