Inventory Planning for a Modular Product Family

• Published in: Production and operations management 2015-07, Vol.24 (7), p.1033-1053
• Main Authors: Paul, Anand, Tan, Yinliang (Ricky), Vakharia, Asoo J.
• Format: Article
• Language: English
• Subjects: Decision making models; Demand; fill rates; Inventory management; inventory planning; modular products; Operations management; Studies
• ISSN: 1059-1478; 1937-5956
• DOI: 10.1111/poms.12370

This paper is motivated by observing that an increasing number of firms are offering modular products assembled with multiple option choices for the consumer. Starting with the PC offerings by Dell which allowed (and still allows) users to configure their product by choosing among multiple choices for each option, the current market place seems to have evolved to a make‐to‐stock scenario where Apple offers its IPAD series with multiple models each with a unique storage size, color, and wireless chip technology. The focus of our work is on determining the optimal stocking level of modular end‐products. Our analysis is based on a benchmark model with the aim of maximizing expected profit subject to an aggregate fill rate constraint as well as variant‐specific individual fill rates under a make‐to‐stock setting. To further assess the robustness of our finding, we consider the extensions of correlated market preferences over options, price‐dependent demand, and alternative probability distributions for characterizing uncertainty in market preferences or aggregate demand. Finally we also show how to extend the single period model into a multiple‐period setting. Through extensive computational analysis, we find that more precise estimates of market preferences for various modular options constitute extremely valuable information that goes beyond the usefulness of forecasts of aggregate market demand. From a practical perspective, this might be indicative of another classic marketing‐operations trade‐off. Offering more options for consumers would be preferred by marketing managers since this would reach more consumers and hence, enhance product sales. On the other hand, the ability to obtaining greater forecast accuracy would decline when the number of options increase. Hence, from an operational perspective, it would be preferred to limit option choices (so that better forecasts can be obtained) since this would lead to lower stocking costs and hence, higher profits.