Group‐level information decay and inventory inspection: An empirical‐analytical approach

• Published in: Journal of operations management Vol. 68; no. 2; pp. 130 - 152
• Main Authors: Chuang, Howard Hao‐Chun, Oliva, Rogelio, Kumar, Subodha
• Format: Journal Article
• Language: English
• Published: Boston, USA Wiley Periodicals, Inc 01.03.2022; Wiley Subscription Services, Inc
• Subjects: Cost analysis; cycle counting; Decay; Empirical analysis; group inspection policy; information decay; Inspection; Inventory; inventory record inaccuracy; multi‐methodological research; Optimization; Optimization techniques; retail operations; Retailing; survival analysis
• ISSN: 0272-6963; 1873-1317
• DOI: 10.1002/joom.1167

Despite numerous optimization techniques proposed to address inventory record inaccuracy (IRI), there continues to be a lack of empirical appreciation of IRI, nor do existing item‐level inspection policies accommodate managerial needs for allocating inspection effort to groups of items. We address these issues by collaborating with a retailer to collect archival data and develop statistical estimation and practical optimization techniques that address IRI. We propose a continuous‐time model to explain group‐level information decay and, based on the model's structure, develop an inspection policy that satisfies managerial needs for allocating inspection effort to a group of items. After analyzing model behavior and quantifying unobservable cost factors using information readily available, we test the efficacy of the proposed policy within the retailing environment and find it to outperform both the retailer's inspection routines and a well‐established industry benchmark. Finally, we illustrate how to optimize grouping and inspection decisions jointly and show potential benefits from optimized group formation. Our empirically grounded modeling effort yields transferable methods and generalizable insights for improving inventory integrity. Highlights • We estimate the unseen hazard rate of inventory information degradation using correction records. • We develop a data‐driven and easy‐to‐implement inventory inspection policy for a group of items. • We show the proposed policy outperforms retailer's current practice and industry benchmark.