Elliot Rabinovich

Assessing the effects of operational processes and information systems on inventory performance

Over the past decade, manufacturing firms have implemented enterprise-wide information systems in order to improve their performance. In many cases, the achieved results have been much less than anticipated. Previous research suggests that in order to realize maximum performance improvements it is necessary to access timely and informative customer demand data. While enterprise-wide information systems may contribute to the realization of this objective, operational processes, such as inventory management and mass customization may also be effective in improving the firms inventory performance. This paper develops a structural equation model that evaluates the impact of enterprise-wide information systems, along with four operational processes, on inventory performance. The results indicate that the use of operational processes, especially just-in-time (JIT), rather than the use of enterprise-wide information systems, leads to improved performance as measured by inventory speculation, lead times, and turnover ratios.

What is the “right” inventory management approach for a purchased item?

Purpose – How should a firm decide which of four choices – i.e. inventory speculation, inventory postponement, inventory consignment, and reverse inventory consignment – is most appropriate to adopt for a given purchased item in a particular context? This paper seeks to identify and explain the critical factors that drive this decision.Design/methodology/approach – By conducting a review of relevant literature and deriving anecdotal observations from four case studies.Findings – This decision is influenced by three factors – customer demand or usage requirements, nature of the supply line and bargaining power of a firm relative to the supplier.Research limitations/implications – From the perspective of science, the conduct of both empirical research to augment the reported anecdotal evidence and conceptual research along a number of directions (e.g. to juxtapose the research findings in existing theories, to examine variations of the four “pure” inventory management approaches, or to consider the vantage ...

POSTPONEMENT EFFECTS ON INVENTORY PERFORMANCE AND THE IMPACT OF INFORMATION SYSTEMS.

Time and form postponement are key to the cost-effective fulfillment of product demand when market volatility is high. Using data from U.S. manufacturing firms, an empirical model suggests that the implementation of time postponement supports the adoption of form postponement. In turn, form postponement is linked to the improvement of inventory management performance. The model also shows that inventory management information systems are coupled with improvements in performance by facilitating a closer alignment between time and form postponement.

Enterprise-wide adoption patterns of inventory management practices and information systems

The adoption of inventory management practices and their relationship with the enterprise-wide adoption of information systems are empirically evaluated. Univariate and multivariate statistical analyses of survey responses indicate a strong enterprise-wide emphasis on materials requirements planning (MRP)- and just-in-time (JIT)-based inventory management systems as well as a continued relevancy of the economic order quantity model. Furthermore, the investigation suggests a positive correspondence between enterprise-wide information systems and MRP adoption and a negative correspondence between enterprise-wide information systems and JIT adoption.

Customer Use of Virtual Channels in Multichannel Services: Does Type of Activity Matter?

Many firms have recently adopted virtual channels, based most notably on the Internet and the phone, to complement the delivery of services to their customers by their existing physical facilities. The success of such multichannel (MC) strategies relies on the alignment of service design decisions—namely those concerning the allocation of service activities to virtual channels—with customers’ MC behavior. Although prior studies have looked at the intensity with which customers use virtual channels, they have not addressed virtual channel use for different types of service activities. In our study, we investigate whether customers’ use of virtual channels for MC services varies with the type of service activities they engage in, and if so, in what way. In doing so, we address two objectives. First, we investigate the impact of accessibility to the physical channel on the degree of use of virtual channels (Internet and phone, aggregated) for different types of activities. Second, we look at channel preferences (Internet vs. phone) for different types of activities when customers do resort to virtual channels to conduct activities. To address our objectives, we develop and test hypotheses regarding customers’ use of virtual channels based on the match between activity attributes (complexity and volume) and channel attributes (access efficiency, interface efficiency, interface richness). Using data from a MC bank, we find that the impact of accessibility to physical channels (specifically, customer distance) on customers’ use of virtual channels, as well as the relative use of Internet versus phone, depend on the type of activities.

Last-mile Supply Network Distribution in Omnichannel Retailing: A Configuration-Based Typology

This monograph develops a configuration-based typology that describes last-mile supply network (LMSN) distribution configurations in omni-channel retailing. The goal is to integrate relevant terms that have been used disjointedly with fragmented bodies of theory to describe the various forms of LMSN. A review of the academic and practice literature was conducted and complemented with secondary observations to identify the key configuration dimensions. Established guidelines for typology and theory building (e.g., Doty and Glick, 1994) were employed to develop the typology. The proposed typology comprises four ideal archetypes: Simple LMSN, Hyperlocal LMSN, One-Stop LMSN, and Protean LMSN. The four are described along the dimensions of: (1) network structure, (2) network flow, (3) relationship governance, and (4) service architecture. Referred to here as the SHOP typology in LMSN, each archetype is associated with one or more core logistics capabilities elaborated upon in a capability profile for each archetype. The typology identifies useful LMSN patterns and enables scholars to develop models and theories based on the four configurations and relate their findings to a specific configuration or across them, rather than expanding efforts on separate and unconnected studies. Notwithstanding, it incorporates elements of the omni-channel context that updates the previous “chain†centric typology developed by Boyer and Hult (2005), and serves as a stepping-stone toward improved insights on what drives, facilitates and inhibits “fit†potential of LMSN configuration. The research output enhances managers’ understanding of the various forms of LMSN distribution configurations and assists in the identification of possible routes to establish configuration footprints across different LMSN forms to support their omni-channel retailing strategies. Further study can identify the critical contingency factors influencing configuration choices.

Inventory disclosure in online retailing

Unlike in a traditional store environment where inventory is directly visible to customers, Internet retailers can selectively choose how to divulge inventory level information to customers. For example, when viewing a particular item page, online shoppers may either see merely “in stock” or a specific inventory level. By choosing the appropriate inventory-level cue to display, a retailer can selectively signal stock-out risk to customers. In this paper, we analyze the optimal structure of an online retailers inventory disclosure and pricing policy, in a two-period setting—a regular selling period followed by a clearance period. Consumers may potentially face uncertainty regarding the firms inventory level, as well as the overall market demand. We show that there exists an inventory level threshold below which a retailer should optimally disclose inventory, and above which masking (i.e., showing only “in stock”) is optimal.Even though the optimal price decreases in the stock level, we show that equilibrium full-price sales may increase or decrease, highlighting the non-intuitive consumer behavior implications of selective inventory disclosure. The optimal “threshold-type” inventory disclosure that we derive reflects the practice of several prominent online retailers selling fashion products, and is new to the literature as prior models of inventory sharing invoked assumptions that led to either consistent disclosure or consistent masking to be optimal. We also extend the model to consider a stochastic market size, and thus highlight that the threshold policy structure continues to hold with demand uncertainty.

Retail Intermediation and Local Foods

Direct sales of local foods, for instance through farmers markets, have reached a plateau, while intermediated sales (through more traditional retailers) are still rising rapidly. We provide an explanation for the growth of intermediated local foods based on the observation that consumers prefer to buy local foods as part of a shopping basket that includes both local and non‐local items. We test our hypothesis using data from a natural experiment conducted by Relay Foods, an online retailer based in the U.S. state of Virginia, and a multi‐variate logit model of shopping‐basket demand. We find significant complementarities among items in a sample shopping basket that are not reflected in estimates from a discrete‐choice model of category demand. Estimates of a structural pricing‐and‐local‐content model reveal important incentives for retailers to offer local foods as part of a broader selection of grocery items.

Omnichannel Retailing as a Balancing Act between In-Store and Home Fulfillment

Problem Definition: Omnichannel retailers can fulfill online orders by allowing consumers to have their orders delivered from the retailers’ stores to the consumers’ homes (home fulfillment) or collect the orders at the stores (in-store fulfillment). Because the latter option does not involve costly last-mile deliveries to consumers’ homes, retailers have sought to make it more attractive by waiving the associated in-store fulfillment fees indiscriminately to consumers. We analyze the economic value of this strategy and compare it with that of a strategy we develop based on the use of targeted incentives to induce individual consumers to switch to in-store fulfillment after initially choosing home fulfillment for their orders. Academic/Practical Relevance: We contribute an analysis of the economic value of price incentives involving home and in-store fulfillment options in omnichannel retailing. In so doing, we isolate determinants of consumer preferences between these options and show how retailers can use such preferences to develop pricing strategies to improve the economic value of their fulfillment services. Methodology: We use field data to evaluate the pricing strategies based on econometric models as well as machine-learning and routing algorithms. Results: While the removal of in-store fulfillment fees increases revenues from additional demand, these gains are much lower than the losses caused by (1) the revenue lost from the fees no longer collected and (2) the additional operating costs at the stores necessary to fulfill a higher volume of orders. Conversely, the costs from using targeted incentives are lower (almost 3 times) than the savings in last-mile delivery costs. Managerial Implications: To promote in-store fulfillment as an alternative to home fulfillment, retailers must go beyond incentive strategies that reward consumers indiscriminately, regardless of the value they attach to these fulfillment services. We show how retailers can develop dynamic pricing solutions to generate value in these services.

An Empirical Analysis of How Inventory Disclosure Affects Online Retail Sales

Online retailers employ varied inventory disclosure tactics ranging from full disclosure at all times to showing mere stock availability cues, which begs the questions “what is the best inventory disclosure strategy” and “can inventory disclosure influence demand at all?” Past academic research in operations and marketing has not reached a consensus, as some researchers advocated that inventory can stimulate demand by indicating popularity or availability, whereas other papers suggest that in some retail settings, low inventory signals scarcity and should stimulate more imminent sales. In this paper, we try to establish a link between inventory and sales empirically by using a large data set that comprises a wide array of products sold at Amazon.com. Controlling for price effects on sales rates and taking into account endogeneity among disclosed inventory levels, prices and sales rates, we show that low disclosed inventory levels stimulate sales. Moreover, we isolate two distinct mechanisms that underlie our main finding, i.e., the frequency of product sales and the magnitude of sales, by which reductions in disclosed inventory and prices positively influence sales rates.