Abstract: Manufacturing is the lifeblood and foundation of China's economic development, how to promote the transformation and upgrading of manufacturing industry in the era of digital economy is an important issue to achieve high-quality development. At present, China's manufacturing industry is in the development stage from traditional manufacturing industry to high-tech manufacturing industry and then to intelligent manufacturing industry, but it is facing the problems of core technology stuck and difficult transformation and upgrading in the process of development. At the end of 2017, “The Guidance on Deepening the "Internet + Advanced Manufacturing" and Developing the Industrial Internet” was issued, and China's manufacturing industry entered the new stage. The main feature of the policy is to promote the deep integration of the Internet and the real economy, enhance the industrial Internet support services and industrial development capacity, and its release is not only conducive to the formation of a good pattern of mutual promotion and simultaneous upgrading of the real economy and the network, but also can strongly promote the modernization of China's economy. This paper empirically investigates the effect and mechanism of industrial Internet industrial policy on the digital transformation of manufacturing enterprises based on a panel data of A-share manufacturing listed companies from 2012-2021 using the double difference propensity score matching method (PSM-DID) with the release of the policy as a quasi-natural experiment, and analyzes the heterogeneity of the effect. The results show the following results: Firstly, industrial Internet industrial policy significantly promotes the digital transformation of manufacturing enterprises, and the above findings still hold after robustness tests. Secondly, financing constraints play a mediating role in the process of industrial policy promoting digital transformation of enterprises. The support of industrial policy can alleviate the financing constraints of enterprises and provide sufficient financial support for transformation and upgrading thus promoting digital transformation of enterprises. Thirdly, the role of industrial policy in promoting the digital transformation of enterprises is influenced by the degree of market competition, the level of enterprise innovation and the life cycle of enterprises. Specifically, the promotion effect is more significant for enterprises with a high degree of market competition, which are more active in transformation and upgrading in order to maintain their competitive position and increase their market share; the promotion effect is more significant for enterprises with a high level of innovation, which have more advanced infrastructure and talent team and can provide technical and personnel conditions for digital transformation; the promotion effect is more significant for enterprises in the start-up and mature stages, which have flexible management teams, more room for development, and stronger willingness and risk-taking ability to transform, so that they can fully seize the policy opportunities to realize digital transformation. Most of the existing studies on digital transformation focus on the micro consequences of digital transformation, and their conclusions indicate that digital transformation has a positive impact on productivity and economic development, but there are fewer studies on the impact factors of digital transformation. Compared with previous similar studies, the possible contributions and innovations of this paper are reflected in the following three aspects: firstly, using the quasi-natural experiment of the release of the policy as an entry point to evaluate the impact of macro industrial policies on the digital transformation of enterprises from the perspective of micro enterprises, which updates the perspective of industrial policy evaluation; secondly, through multiple heterogeneity analysis, the differences of different types of enterprises affected by industrial policies are examined, which provides references for government; thirdly, the empirical analysis using the difference-in-difference propensity score matching method (PSM-DID) effectively addresses the endogeneity problem of sample selection and makes the research results more convincing.