In recent years, many automation technologies and equipment are applied for new injection molding systems for Industry 4.0. However, a lot of the existed injection machines or systems are not ready for automation yet. Before automatic manufacturing, how to retain good quality is one of the crucial factors in injection molding. In this study, we have proposed a feasible system to study the deviation difference between numerical simulation and experiment in injection molding product development. We also developed a decoupled procedure to analyze the internal mechanism of the deviation between numerical simulation and experiment. To verify our concept, we have performed the decoupled procedures on the proposed round disc system based on the same external driving force changes. Results showed that the internal driving force of shrinkage change is almost the same for both numerical simulation and experiment study. The main reason for the deviation difference is due to the response lag of the injection speed in the real experiment. The deviation difference between numerical simulation and experiment is reduced by 36% after corrected the packing pressure. Moreover, after calibrated the injection response, the deviation difference between numerical simulation and experiment has been reduced by 56%.