It is well known that the greatest development opportunity for China’s banking industry lies in retail banking. Retail banking has become a critical component of banking business transformation due to relatively weak economic cycle fluctuations, lower operational risks, lower personal loan non-performing rates, and increased contribution to operations. Industrial and Commercial Bank of China (ICBC), leveraging its extensive payroll and branch network, once proposed the strategic goal of becoming “China’s top retail bank.” Likewise, Construction Bank aimed for “top-tier retail banking,” and China Merchants Bank positioned itself as the “king of retail.” Given the relatively low capital utilization of retail banking and tightening capital regulatory requirements, commercial banks have successively introduced concepts such as retail digital transformation, large retail segments, and customer operations deepening. The competition in the retail banking sector has intensified, primarily manifesting in price competition, product competition, channel competition, as well as talent and customer competition. With the opening of financial markets and the entry of foreign and private banks, the number of domestic and international banks has shown a slight increase. According to data released by the China Banking and Insurance Regulatory Commission: as of the end of December 2018, there were 4588 banking financial institution legal entities in China; by the end of June 2019, this number had increased to 4597; and by the end of June 2021, it had risen to 4608. The more banks there are, the greater the competitive pressure in retail banking. Under circumstances of product homogenization and intensified competition in the banking industry, precision marketing becomes crucial. Banks can use artificial intelligence algorithms to identify which customers provide greater value returns, thus gaining a larger market share and enhancing competitiveness under equal conditions. The key to retail banking is customer operations, and the key to customer operations is understanding customers. However, this is not an easy task. On the one hand, the sheer number of retail customers makes one-on-one marketing via traditional methods labor-intensive, inefficient, and minimally effective. On the other hand, many clients’ assets are dispersed across multiple banks, securities firms, and insurance companies. Viewing from the perspective of a single bank may not capture the full picture—a “regular client” at one bank may be a high-value client at another. Hence, we need an intelligent, comprehensive data model to identify potential high-value clients.

Retail Potential High-value Customer Identification: Graph Neural Network Technology

章节
发表日期：

2025年

查看全文

Liyu Shao,
Qin Chen,
Min He

摘要

It is well known that the greatest development opportunity for China’s banking industry lies in retail banking. Retail banking has become a critical component of banking business transformation due to relatively weak economic cycle fluctuations, lower operational risks, lower personal loan non-performing rates, and increased contribution to operations. Industrial and Commercial Bank of China (ICBC), leveraging its extensive payroll and branch network, once proposed the strategic goal of becoming “China’s top retail bank.” Likewise, Construction Bank aimed for “top-tier retail banking,” and China Merchants Bank positioned itself as the “king of retail.” Given the relatively low capital utilization of retail banking and tightening capital regulatory requirements, commercial banks have successively introduced concepts such as retail digital transformation, large retail segments, and customer operations deepening. The competition in the retail banking sector has intensified, primarily manifesting in price competition, product competition, channel competition, as well as talent and customer competition. With the opening of financial markets and the entry of foreign and private banks, the number of domestic and international banks has shown a slight increase. According to data released by the China Banking and Insurance Regulatory Commission: as of the end of December 2018, there were 4588 banking financial institution legal entities in China; by the end of June 2019, this number had increased to 4597; and by the end of June 2021, it had risen to 4608. The more banks there are, the greater the competitive pressure in retail banking. Under circumstances of product homogenization and intensified competition in the banking industry, precision marketing becomes crucial. Banks can use artificial intelligence algorithms to identify which customers provide greater value returns, thus gaining a larger market share and enhancing competitiveness under equal conditions. The key to retail banking is customer operations, and the key to customer operations is understanding customers. However, this is not an easy task. On the one hand, the sheer number of retail customers makes one-on-one marketing via traditional methods labor-intensive, inefficient, and minimally effective. On the other hand, many clients’ assets are dispersed across multiple banks, securities firms, and insurance companies. Viewing from the perspective of a single bank may not capture the full picture—a “regular client” at one bank may be a high-value client at another. Hence, we need an intelligent, comprehensive data model to identify potential high-value clients.