Channel structures and subscription strategies for AI-driven logistics data products

Abstract

Motivated by the application of large models in artificial intelligence (AI), this paper proposes a new business model for AI-driven data product transactions in the freight market. We develop a game-theoretic model for the logistics data supply chain comprising a logistics data provider and a logistics data integrator. Observing the opportunity for the logistics data provider to directly sell AI-driven data products to consumers and supply data sets to the logistics data integrator, we explore two channel structures: a single-channel structure and a dual-channel structure. Furthermore, the logistics data provider can choose whether or not to subscribe to the value-added services provided by Cyber–Physical Internet (CPI), which enhance data product quality but also incur additional costs. This study presents the following results. First, our findings debunk the prevailing belief about product quality strategy that improving data product quality instead impairs the profit when targeting a high licensing rate and a large number of affluent consumers. Second, a dual-channel structure is only viable if the licensing rate is sufficiently high or the market is dominated by budget-conscious consumers, otherwise a single-channel structure is a superior choice. Third, subscribing to the value-added services provided by CPI, even when free, may not benefit the logistics data provider due to the spillover effect in a dual-channel structure. Managerial implications enable logistics data providers to achieve greater economic efficiency under various market conditions by adopting suitable channel structures and leveraging value-added services and pricing tools, thereby promoting AI-driven data product transactions.