Should live broadcasting platforms adopt artificial intelligence? A sales effort perspective

We analytically investigate a supply chain (SC) that is composed of a manufacturer and a live broadcasting platform, and examine whether the latter should adopt artificial intelligence (AI) considering sales effort. We consider several important factors that affect the SC partners' decision-making including live broadcasting power, consumer expectations of the product, and unfit probability that the consumer is unsatisfied with the bought product. Specifically, the "live broadcasting power" refers to the power of influencers' personal influence and fans group to enhance product sales. The results are as follows: First, the optimal production quantity of the offline channel (platform) exhibits a positive (negative) correlation with the retail price of the offline channel in the agency mode with AI. Nevertheless, in the resale mode, the retail price of the offline channel has no influence on the two channels' optimal production quantities. Second, with low marginal cost of adopting AI, the live broadcasting platform should (not) adopt AI under high (low) live broadcasting power. With high marginal cost of adopting AI, the live broadcasting platform should (not) adopt AI under low or high (moderate) live broadcasting power. In addition, the manufacturer without AI should select the agency mode (resale mode) under high (low) live broadcasting power, while the manufacturer with AI should always collaborate with the live broadcasting platform implementing the agency mode. Finally, the agency and resale modes can achieve coordination between the two firms. We also consider the partial unfit probability, hybrid mode, and "webrooming" behavior to extend our study, and numerically demonstrate our analytical findings' robustness.