Advancing algorithmic bias management capabilities in AI-driven marketing analytics research
研究提出工业市场中算法偏差管理能力的框架，包含数据、模型和部署三个主要维度及九个子维度，通过文献综述、主题分析和两轮调查验证，对提升客户权益有重要意义。
Algorithms in the age of artificial intelligence (AI) constantly transform customer behaviour, marketing programs, and marketing strategies in industrial markets. However, algorithms often fail to perform as expected due to various data, model, and market biases. Motivated by this challenge, this study presents a framework of algorithmic bias management capabilities for industrial markets that contribute to customer equity in terms of value, brand and relationship equity. Drawing on the dynamic capability theory, this study fills this gap by conducting a literature review, thematic analysis, and two rounds of surveys (n=200 analytics professionals and n=200 business customers) in the financial service industry in Australia. The findings show that algorithmic bias management capability consists of three primary dimensions (data, model, and deployment capabilities) and nine subdimensions. These findings have important implications for scholars and managers interested in developing algorithmic bias management capabilities to influence customer equity in industrial markets.