Scholarlink

<p>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.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The Challenge of Digital Transformation for Small and Medium-Sized Enterprises in Japan

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月14日

查看全文
  • Kazunori Minetaki,
  • Teruyuki Bunno,
  • Hiroki Idota

摘要

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.