Abstract:    With the rapid development of the Internet, 5G and other new generations of information technology, enterprises need to cross-domain organization boundaries to capture innovative opportunities, inspire innovative inspiration and cultivate new skills so as to improve the innovative ability of enterprises through R&D search behavior. However, the existing researches ignore the motivation for the R&D search. In fact, taking what kind of R&D search strategy, whether or not to carry out R&D search and when to conduct R&D search are the key strategic issues in the process of enterprise innovation. Therefore, whether it is caused by problem making or predicting is very important for R&D search behavior. Based on the relevant theories of organizational behavior (OB), with the foundation of analyzing the differences between performance feedback and judging the situation, the alliance routines are incorporated into the research model to study the strategic impacts of performance feedback, judging the situation on R&D search behavior under the moderate of alliance routines. Based on the data from the Shanghai Stock Exchange website and questionnaire, this paper selects 295 enterprises in the manufacturing industry, construction industry and information transmission industry as samples, and analyzes the strategy selection of enterprises′ R&D search behavior combined with the research methods of hierarchical regression.The results show that when the performance feedback is lower than the expected value, enterprises will strengthen R&D search; when the performance feedback is higher than the expected value, enterprises will reduce R&D search intensity; when the decision-maker expects the future performance to be higher than the future expectations, companies will reduce the R&D search. When the performance feedback is lower than the expected value and the expected future performance is better than the future performance, alliance routines play a negative moderating role in it. But when the performance feedback is higher than the expected value, alliance routines strengthen the R&D search behavior of enterprises. Through the study of this paper, we also find that the changes in short-term performance have an important impact on the future R&D search behavior of enterprises.This research results fill the gaps in the theoretical part of enterprise R&D search. The existing researches on R&D search mainly focus on the classification of R&D search, and then study their impact on innovative performance. Combining with the relevant research of rational person, this paper proposes a variable of judging the situation to further explain the R&D search behavior of the enterprise based on the performance feedback research. It is believed that the gap between the expectation of the shareholders for future corporate performance and the expected reality will directly affect the R&D search intensity of enterprises. When the expectations are lower than the future performance, enterprises will strengthen the R&D search; At the same time, it is found that the comparison between the future satisfaction performance of the decision maker and the future expected performance will also have an impact on the company′s R&D search behavior. Therefore, the existing research on search theory is expanded.The results also show that alliance routines have different degrees of influence on R&D search behavior through their gradual and fixed behavior pattern. Alliance routines play a negative moderating role when the performance feedback is lower than the expected value and the expected future value is better than the future performance. However, when the performance feedback is higher than the expected value, alliance routines strengthen enterprises′ R&D search behavior. This research has important practical significance for the selection behavior of R&D search intensity. In the strategic choice of R&D search, enterprise decision-makers can guide organizational change or innovation through appropriate target levels and expectation value. Not only can they respond to performance feedback, but they can also take proactive action to make a full, detailed and scientific prediction of the future and make careful preparation in advance.

Key words: performance feedback, judging situation, R&D search behavior, alliance routine, strategy selection