Improving Order Fulfillment Performance through Integrated Inventory Management in a Multi‐Item Finished Goods System

• Author: Haitao Li,Liu Yang,James F. Campbell
• Date: 01 March 2020
• Published date: 01 March 2020
• DOI: http://doi.org/10.1111/jbl.12227

Improving Order Fulfillment Performance through Integrated

Inventory Management in a Multi-Item Finished Goods System

Liu Yang

1

, Haitao Li

2

, and James F. Campbell

2

1

Sam Houston State University

2

University of Missouri –St. Louis

Effective inventory management is fundamental to order fulfillment excellence and supply chain success. In this paper, we develop a strate-

gic inventory management decision tool that integrates inventory classification and inventory control policy decisions for maximizing order

fulfillment performance, while accounting for a constraint on inventory budget and the profit expectation of a firm. This inventory solution tool

provides critical enhancements to current inventory planning software, which is developed upon the traditional inventory classification scheme

and where practitioners have to balance service levels and safety stock decisions through trial-and-error. The model allows firms to assess

whether the current inventory performance is Pareto optimal, quantify the trade-offs between various performance measures, and identify the

right inventory level according to the firms’strategic goals. In computational results, we demonstrate the trade-off and positive relationships

between key item- and order-based inventory performance measures and short-term profitability under different levels of inventory budget in a

multi-item finished goods inventory system.

Keywords: order fulfillment optimization; inventory classification; order fill rate; trade-off of profit and service level; mixed-integer

programming

INTRODUCTION

A rich stream of interdisciplinary research on logistics and market-

ing has shown that order fulfillment performance is a significant

determinant of customer satisfaction and loyalty, directly affecting

customer future purchasing behavior and firm revenue (e.g.,

Daugherty et al. 1998; Stank et al. 2003; Davis-Sramek et al.

2010; Rao et al. 2011; Griffis et al. 2012). A fundamental element

dictating order fulfillment performance is inventory control poli-

cies. While single-item inventory control has been extensively

studied, there is little research available to guide companies on

establishing inventory policies for order fulfillment improvement

in a multi-item finished goods inventory system. The modern

inventory planning software packages are developed upon the tra-

ditional inventory management approach and concern the service

level of stock keeping unit (SKU) rather than customer orders.

Typically, companies group the large variety of items into a few

classes according to the Pareto principle or multicriteria inventory

classification (MCIC) scheme and then assign each class a subjec-

tively determined cycle service level (CSL; i.e., the expected prob-

ability of not having a stock-out during a replenishment cycle;

Mohammaditabar et al., 2012; van Kampen et al. 2012; Teunter

et al. 2017). For example, Microsoft prioritizes its hardware prod-

ucts into A, B, C, and D classes based on item revenue, life cycle

status, profit contributions, and marketing factors and then sets

CSL targets for each class (Neale and Willems 2009). With the

help of inventory planning software, required inventory level is

computed and reviewed by the management. If the inventory level

is acceptable, the process is completed; otherwise, the inventory

planner needs to adjust the CSL of each class until the inventory

level is within the budget. Figure 1 depicts this process. It requires

trial-and-error during inventory planning to balance service level

and safety stock. Moreover, the entire procedure is performed

independently of order fulfillment measures; it takes no considera-

tion of order structure, and it is unclear how the inventory classifi-

cation and the class-based service levels may impact order

fulfillment. It has long been argued that the classification criterion,

the number of classes, the determination of the cutoff value

between classes, and the service level of each class rely more on

managerial judgment than quantitative analysis (e.g., Viswanathan

and Bhatnagar 2005; Stanford and Martin 2007; Teunter et al.

2010; Lajili et al. 2012). A review of the inventory literature has

not identified any work that simultaneously optimizes inventory

classification and service level (and the corresponding safety

stock) decisions to maximize order fulfillment measures.

In addition, supply chain is a complex network, requiring com-

panies to continuously weigh and balance key trade-offs in inven-

tory management. The leading-edge inventory planning software

offers trial-and-error exploration of trade-offs and alternatives, but

is limited in two aspects (Figure 2). First, it is done at individual

SKUs or SKU-class levels instead of a system-wide approach; sec-

ond, only a small number of factors are considered in decision

making, while the important business aspects such as inventory

budget, short-term profit requirement, order criticality, and order-

varying pricing are not accounted for. To fill these gaps, we

develop a strategic inventory management decision tool in concert

with a food product manufacturer to address the order-fulfillment-

optimized inventory classification and safety stock decisions, while

meeting the profit expectation of a firm and subject to an inventory

budget constraint. The solution takes a holistic, systems approach

to the inventory management and offers critical enhancements to

the current inventory planning software packages.

Consider a particular order (order k) that consists of multiple

items. The fill rate for this order, denoted as R

k

,isdefined as the

fraction of items in order kthat are filled from on-hand

Corresponding author:

Liu Yang, College of Business Administration, Sam Houston State

University 1905 University Ave, Huntsville, TX 77340, USA;

E-mail: LYang@shsu.edu

Journal of Business Logistics, 2020, 41(1): 54–66 doi: 10.1111/jbl.12227

© 2019 Council of Supply Chain Management Professionals