Abstract:In order to answer the question that “Why can’t the R&D talents introduced by enterprises stay”, this study is established in job seekers’ needs, which uses Python big data crawler technology to craw for 15 days big data in comments on job seekers for R&D positions on major recruitment websites and related job forums in China and is based on P-O fit theory. This study summarizes the structural dimension and quantitative observation method of enterprise recruitment information quality via Nvivo 15.0 software. Besides, we discover that the quantity of enterprise recruitment will adjust delay the feeling and experience of new employees after entering the company though influence their job nesting cognitive accuracy, and continue to adjust the impact of their willingness to be retained after they produce job embeddedness. According to the result of study, we suggest that related enterprises improve the process of previous employment,explode related employment information continuously,improve the organization atmosphere and environment mechanism and help new employees enhance retention intention rapidly during the talent introduction in order to contain the brain drain which is caused by new employee expectation gap from the origin.