The invention discloses a big data intelligent marketing system and a method thereof. Through acquiring enterprise related data comprehensively and systematically, a machine learning algorithm and a model are utilized to extract enterprise salient features including risk features, so that accurate portraits can be made for enterprises from multiple dimensions. A plurality of association relationships between the enterprises are utilized to find out an association path between the user enterprise and the target enterprise. Optimized marketing distribution is carried out by combining target enterprise characteristics and various information of sales personnel or sales channels, and marketing effect data are collected to continuously optimize a distribution mechanism. Therefore, users can automatically achieve the purposes of accurate matching of target enterprises, reasonable suggestions of marketing schemes, distribution optimization, improvement of the marketing conversion rate of thetarget enterprises and the like through a computer in a datamation mode.