Unlearn Success or Failure Beliefs?: How Do Big Data Analytic Capabilities Affect the Incumbents' Business Model Innovation in Deep Uncertainty

Research investigating the underlying mechanisms and boundary conditions under which Big Data analytic capabilities (BDACs) influence business model innovation (BMI) in incumbents remains largely underdeveloped. Drawing on the dynamic capabilities view (DCV), we developed a moderated multimediation model in which unlearning success beliefs and unlearning failure beliefs were theorized as the different mechanisms underlining why incumbents are more likely to engage in BMI under the influence of BDACs. We further proposed that deep uncertainty is an important boundary condition that affects such a relationship. Multisource data from a multiwave survey was analyzed using structural equation modeling to test the theoretical framework. The results indicated that BDACs positively affect incumbents' BMI through not only unlearning success beliefs but also unlearning failure beliefs. Furthermore, the results provided evidence for that deep uncertainty positively moderates the mediation of unlearning success beliefs. Notably, although the moderating effect of deep uncertainty on the mediation of unlearned failure beliefs is negative, it is insignificant. Our study contributes theoretically to the research on BDACs, organizational unlearning, BMI, and DCV, while practical implications are also discussed.