Navigating digital transformation for inventory leanness through product and supply chain strategy

Purpose This study examines why digital transformation has a non-monotonic effect on inventory leanness and how product diversification and supply-chain concentration condition this relationship. Design/methodology/approach The empirical analysis uses an unbalanced panel of 2,533 Chinese listed manufacturing firms and 19,299 firm-year observations from 2010 to 2023. Firm and year fixed effects are absorbed and conventional standard errors are reported in the primary models. Following Haans et al. (2016), the quadratic specification is treated as a parsimonious empirical approximation to a theory-derived marginal-effect pattern; the analysis tests endpoint slopes, the Lind–Mehlum joint U test, and Fieller confidence intervals; and evaluates conditional curvature, turning-point movement and shape-flip in the moderated models. Findings The findings reveal a U-shaped relationship between digital transformation and inventory leanness. Product diversification strengthens the curvature of this relationship, and this curvature-enhancing effect is further amplified by supplier concentration but weakened by customer concentration. Conditional analyses indicate that the complete U-shaped pattern depends on specific configurations, with limited shape reversals occurring only at the distributional tails. Practical implications Managers should distinguish exploration-related learning friction from the exploitation of retained digital routines. Product diversification raises the information-processing burden during early transformation but can create stronger inventory benefits after digital routines mature. Deep supplier coordination can reinforce these benefits through relation-specific digital assets, whereas dependence on a small number of powerful customers can limit portfolio-wide inventory optimization. Originality/value The study contributes by identifying two countervailing latent mechanisms – exploration-related learning friction and the exploitation of retained digital routines – that combine additively to generate a U-shaped relationship. It further shows how product complexity alters the curvature of this relationship and how supplier and customer concentration provide asymmetric boundary conditions. The analysis responds to methodological concerns in nonlinear research by separating curvature change, turning-point movement and shape-flip.