This study aims to investigate the factors affecting the implementation of Digital Transformation (DT) and verify the causal relationship between it and new product/service innovation. The data analysis is based on the “Questionnaire on Innovation Activities in Small and Mid-Size Manufacturing Industries.” 3000 manufacturing industry firms were selected from members registered with the Osaka Bureau of Industry, and 433 valid responses were obtained. The analysis methods used were SHAP (SHapley Additive exPlanations), for factor analysis on DT, and CATE (Conditional Average of Treatment Effect), one of the causal inference methods of machine learning, for the effect of DT on innovation. Three conclusions can be drawn: (1)The purposes of implementing DT, which are (a) improving customer loyalty through enhanced online sales and marketing and (b) operational efficiency and productivity by improving business processes using data, were effectively linked with the result. For the former, the contribution was found to be high despite the small number of companies concerned. (2) It was found that adopting AI is interlinked with DT. (3) Machine learning causal inference verified that implementing DT can lead to new product/service innovation.
Minetaki et al. (Sat,) studied this question.