Vendor-managed inventory in fresh-food supply chains
Post, Roel
DOI:
10.33612/diss.130028783
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Post, R. (2020). Vendor-managed inventory in fresh-food supply chains. University of Groningen, SOM research school. https://doi.org/10.33612/diss.130028783
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Introduction
The organization of retail supply chains is a challenging process in which suppliers and retailers collaborate to match supply and demand as closely as possible. Too little supply can lead to lost sales and unhappy consumers, while too much sup-ply leads to a waste of resources and expensive markdowns, in particular for per-ishable products. Aligning all replenishment decisions (i.e., managing production, transportation and inventory) so that supply meets demand in a supply chain is a complex process in and by itself. On top of that, in traditional supply chains, respon-sibility for these decisions are divided over a retailer and its suppliers. Typically, the retailer holds inventory in their own warehouses to fulfill the customer demand and places orders at its suppliers to replenish this inventory. This is known as Retailer Managed Inventory (RMI). The suppliers produce the products and transport the products to the retailer in the ordered quantities. Often, a supplier also holds its own inventory. Coordinating such a process between organizations is demanding, which often causes suppliers and retailers to make decisions that are sub-optimal from the perspective of the supply chain as a whole.
On top of this, the strong competition retailers are facing, both online and offline, has driven them to expand their assortment. To obtain similar revenues, retailers sell a much higher number of products, supplied in increasingly smaller volumes per product. On top of this, the appealing effect of fresh products on consumers has led to a reduction in the average shelf lives of the products, in particular for grocery retailers. In the past, retailers could oversee the replenishment for the limited
num-Inventory Vendor Production planning Transportation planning Inventory retailer Store-demand Production planning Transportation planning Inventory retailer Store-demand Order
Demand and inventory information
Vendor Retailer Retailer Vendor RM I VM I Inventory vendor
Figure 1.1: Schematic overview.
ber of fast-moving products with stable demand. Taking into account the specifics of larger numbers of slow-moving products, however, is no longer possible for a retailer’s planners and account managers.
To counter this challenge, many suppliers (e.g., the Campbell Soup Company) and retailers (e.g., Walmart) have pushed for a different way to organize their re-plenishment processes, originally called efficient consumer response or continuous replenishment, but later became known under the name vendor-managed inventory (VMI). Under VMI, a retailer no longer sends orders to its suppliers, but shares infor-mation on its inventory positions and sales. The supplier then becomes responsible for the decisions regarding inventory in the retailer’s warehouse, and thus effectively manages the entire replenishment process (Figure 1.1). Thanks to the many reports in both academic literature and industry press about successful implementations in which VMI helped to improve demand fulfilment and lower inventories, VMI be-came (and still is) a widely adopted process in all kinds of industries, ranging from grocery, to hardware stores to automotive manufacturers.
Despite its popularity, not everything about VMI has been good news. Next to several success stories (e.g., Waller et al., 1999), there are also many examples where the envisioned performance improvements, i.e., lower inventories and stockouts, did not occur and where VMI was eventually revoked (e.g., Sari, 2008; Dong et al., 2014). In these VMI implementations, the suppliers were not able to make high-quality
re-plenishment decisions causing the anticipated gains for the supply chain to not al-ways materialize. Or, the gains did materialize, but were not equally distributed across the retailer and suppliers causing friction or one of the organizations to suffer. Understanding the outcomes of a VMI implementation is further complicated by the diversity in the contexts in which suppliers operate; even a single supply chain con-sists of many different suppliers that are all unique in the products they ship and the production processes they operate. This may explain why in some situations VMI was only revoked for part of the suppliers involved (e.g., Dong et al., 2014). The in-consistency in outcomes of VMI renders scholars and practitioners unable to answer the seemingly straightforward but important questions: How will the performance of a supply chain change when a retailer implements VMI? And, for which suppliers does VMI hold (the most) potential?
In this thesis, we will explore how introducing VMI affects the replenishment decisions of a supplier and how the supply chain context is influencing these de-cisions. Leaning strongly on inventory and information theory (Cachon, 2004; Ru et al., 2018; Kadiyala et al., 2019), prior studies have shown that VMI holds the po-tential to improve supply chain performance in many different ways. In the next two sections of this introduction, we built upon this literature to characterize VMI as a multi-faceted concept that can appear in many different forms (Fry et al., 2001; Elvander et al., 2007) and consists of different ’components’ that can be used to im-prove the performance of a supply chain. In the third section of this introduction, we will show why the implication of these components have been under-explored in the current empirical literature and how this shaped the current understanding on VMI. In the last section of this introduction, we will outline how this thesis addresses this gap in our understanding, and how new insights in VMI can be used to enhance the performance of fresh food supply chains.
1.1
Control and ownership of inventory in a supply chain
VMI is all about the division of control and ownership of inventory in the supply chain, i.e., which organization makes the inventory decisions and who incurs the (fi-nancial) consequences of those decisions. As explained above, in a traditional supply chain the retailer makes the decisions regarding inventory levels in their warehouse. In VMI, this control of inventory is transferred to the supplier. However, if only
Inventory ownership Invent ory cont rol Supplier Supplier Retailer Retailer Vendor Managed Inventory (Continuous replenishment, Efficient consumer response)
Consignment Inventory Vendor Managed
Inventory with Consignment Inventory
(Scan based trading)
Traditional supply chains
(Retailer Managed Inventory)
Figure 1.2: Traditional, Vendor-Managed Inventory and Consignment Inventory.
the control of inventory is transferred, the supplier decides on the inventory levels for which the retailer incurs the consequences in the form of holding costs or sales price markdowns. This lack of alignment in control and ownership can incentivize a supplier to make inventory decisions in its own favor, rather than acting in the best interest of the supply chain. For example, a supplier can ’push’ inventory to the retailer’s warehouse beyond levels that are necessary to fulfill demand, while the retailer is paying for the excess inventory. Therefore, many VMI implementations also involve a transfer of ownership, often called consignment inventory. Under consignment inventory, the supplier remains owner of the products until they are taking from the inventory position to fulfill customer demand. Figure 1.2 provides a schematic overview of the possible VMI configurations with respect to inventory ownership and control. For a more extensive overview of different VMI configura-tions we refer to Elvander et al. (2007).
The configuration of a supply chain is often closely related to the power relations in this supply chain. In supply chains with a powerful supplier, the supplier is more
likely to be the one that initiates introduction of VMI. Because this supplier has little incentive to also take on ownership of the inventory—it would simply increase its holding costs—the retailer often remains owner of the inventory in its warehouse (see, e.g., Clark and Hammond, 1997; Dong et al., 2014). To ensure a supplier does not make decisions at the expense of the retailer, contractual safeguards in the form of limitations on inventory levels are often implemented (Fry et al., 2001). VMI im-plementations driven by a powerful retailer often do not only involve a transfer of control of inventory, but also a transfer of ownership, due to the retailer’s bargain-ing power and motivation to reduce their inventory holdbargain-ing costs. The transfer of both control and ownership forms a more ’natural’ incentive for suppliers to opti-mize the supply chain because they both make the inventory decisions and incur the consequences of those decisions.
1.2
The theoretical mechanisms of VMI and
consign-ment inventory
VMI has mainly been described as a collaborative effort in which the information transfer component (i.e., where the retailer shares additional demand and inventory data with its suppliers) plays a crucial role. Authors of early empirical studies on VMI even noted that the observed improvements of VMI implementations could have been realized by information sharing alone (Cachon and Fisher, 1997). More recently, Dong et al. (2014) were the first to show empirically that adding the deci-sion transfer component of VMI (i.e., where the suppliers become responsible for re-plenishment frequencies and quantities) adds significant benefits that go beyond the information sharing component alone. In many cases, an inventory position trans-fer component (i.e., where the physical position and ownership of the inventory is moved from the suppliers to the retailer) is part of a VMI implementation as well (Elvander et al., 2007), on top of information sharing and the decision transfer. Fi-nally, the ownership of inventory can be transferred to the supplier as part of a VMI implementation. Below, we will describe the potential of each component of VMI to improve supply chain performance, and summarize these potential improvements in Table 1.1.
VMI component Potential supply chain performance gain
Information transfer The information transfer component of VMI provides
poten-tial to improve supply chain performance if it helps reduc-ing uncertainty in the replenishment process and thereby re-duces the need for buffers. Moreover, it can reduce the bull-whip effect.
Decision transfer The decision transfer component of VMI provides potential
to improve supply chain performance if it enhances the plan-ning of the replenishment processes by enabling more deci-sion freedom or better decideci-sion making.
Inventory position transfer The inventory position transfer component of VMI provides
the potential to improve supply chain performance if it re-duces buffers needed for process uncertainty and improves the response time as a result of keeping inventory further downstream the supply chain.
Ownership transfer The ownership transfer component of VMI provides the
po-tential to improve supply chain performance if it promotes supply chain level profit optimization by aligning decisions and their consequences, enabling competition between sup-pliers, and preventing double marginalization.
Table 1.1: The potential for supply chain improvement of each VMI component.
1.2.1
Information transfer component
It has long been established that information sharing can improve supply chain per-formance by reducing uncertainty (Lee et al., 2000). The information transfer compo-nent of VMI involves sharing customer demand volumes and inventory level infor-mation and sometimes sharing knowledge of inventory management policies. The information transfer component by itself (i.e., without any of the other VMI compo-nents) can improve the planning of those replenishment processes that were already under control of the supplier – being sourcing, production and transportation de-cisions. Customer demand and inventory level information enables a supplier to better estimate the timing and quantity of the retailer’s demand and thereby facili-tates planning its replenishment processes at an earlier point in time. That additional time could enable better synchronization and reduces the need for buffers (Kaipia et al., 2002). Furthermore, sharing information can reduce the bullwhip effect (Lee et al., 2000; Disney and Towill, 2003), since the information potentially facilitates planning at a supply chain level instead of at the level of separate inventory po-sitions. Of course, the above-described improvement potential is only realized if a
supplier proactively uses the information to actually change its planning, production and transportation processes.
1.2.2
Decision transfer component
The decision transfer component of VMI moves the responsibility for replenishment decisions from a retailer to a supplier. Exerting control over the replenishment pro-cesses boils down to deciding the timing and (lot size) quantities for each step in the replenishment process (e.g., production and transportation). Together, these plan-ning decisions form a combined production, inventory, and transportation problem with a number of constraints, such as transport capacity or inventory level limits.
Prior research has shown that the decision transfer component of VMI holds the potential to improve supply chain performance in three ways. First, the decision transfer component could help to improve the quality of the decision rules to solve the combined production, inventory and transportation problem (Clark and Ham-mond, 1997; Disney et al., 2003; Dong et al., 2014). The quality of the decision rules is determined by how close the replenishment decisions attain the inventory lev-els with the highest product availability at minimum costs, given the constraints and available demand estimations. The decision transfer component can lead to better supply chain performance if the quality of the decision rules applied by a supplier under VMI are superior to the decision rules applied by a retailer under retail-managed inventory (RMI). The decision rules applied by a supplier can be superior because a supplier may have access to better information, and/or applies decision rules that over time provide a better trade-off between inventory, missed or backlogged demand, and resource utilization. For example, a supplier may have information on demand and promotions by other retailers it supplies and knows the capacity constraints of its production process, which can be valuable input for inventory decisions (Cetinkaya and Lee, 2000; Cheung and Lee, 2002; Dong et al., 2014).
Second, the decision transfer component may take away additional constraints that could not have been overcome by information sharing alone. Even though a supplier can better anticipate the upcoming demand of a retailer as a result of in-formation sharing, without decision authority over the replenishment process, that supplier is still bound to the delivery timing and quantity decisions set by the
re-tailer. If those inventory levels are too restricting for the supplier, the only thing a supplier can do is to keep—or increase—inventory levels at its own warehouse. With the decision transfer component of VMI, the supplier can also change the inventory at the retailer’s inventory location and change the timing of its deliveries. (Disney et al., 2003)
Third, if a supplier has multiple buyers, that supplier can prioritize production and shipments so that the retailer with the lowest inventory level is served first (Cetinkaya and Lee, 2000). The demand uncertainty of multiple retailers combined is lower than the uncertainty of each retailer individually due to pooling of stochas-tic variance, reducing the supplier’s need for buffers (Lee et al., 2000). In addition, the decision transfer can enable combining deliveries to multiple retailers, which re-duces uncertainty by making a higher delivery frequency possible without loss of transportation efficiency (Waller et al., 1999; Cetinkaya and Lee, 2000; Cheung and Lee, 2002).
1.2.3
Inventory position transfer component
Allowing a supplier to physically position its inventory at a retailer’s inventory lo-cation is often part of VMI (Elvander et al., 2007), but is hardly discussed in the academic literature. If a supplier’s inventory is completely transferred to a retailer’s inventory location, that supplier ships its products directly to the retailer after pro-duction, eliminating the inventory location at the supplier (Ferguson and Ketzen-berg, 2006). It is also possible to re-position only a part of a supplier’s inventory, for example, when a retailer sets minimum or maximum inventory levels (Fry et al., 2001).
Prior research has shown two ways in which this inventory position transfer com-ponent can improve supply chain performance. First, the demand fulfilment of prod-ucts that are already stored in a retailer’s inventory location is no longer impacted by process uncertainties that exist between the supplier’s and retailer’s inventory lo-cations (van der Vlist, 2007). For example, delays in the transport process would no longer lead directly to stock outs. Keeping inventory closer to the location where de-mand occurs reduces the response time and, therefore, reduces the need to build-up buffers that are otherwise needed to cope with possible disturbance in the process over this lead time. Second, by giving the supplier complete control over all
inven-tory in the supply chain, VMI could enable the supplier to plan for “economic man-ufacturing quantities” or “economic order quantities” that make production and/or transport more efficient (Holweg et al., 2005; Glock, 2012).
1.2.4
Ownership transfer component
The inventory position transfer is often implemented together with restrictions on inventory levels (Claassen et al., 2008) or a transfer of inventory ownership, where the retailer pays for the products only when they are shipped to customers (Elvan-der et al., 2007). Authors studying VMI and consignment inventory by means of economic modeling have shown that the advantages of transferring ownership of inventory to the supplier go beyond just preventing opportunistic behavior from the supplier’s side. Even in traditional supply chains where inventory decisions and ownership are aligned, but divided over supplier and retailer, the incentives are of-ten not such that the supply chain as a whole can make the most profit (Cachon, 2004). Double marginalization, i.e., where both the supplier and retailer calculate a profit margin separately, is a well-known cause for fulfilling less demand than would be possibly in an ’efficient’ supply chain. Centralizing both control and ownership of inventory at the supplier enables this supplier to make efficient decisions for the entire supply chain (Bernstein et al., 2006). Moreover, the fact that VMI enables pliers to make their own inventory decisions may increase competition between sup-pliers that serve to the same retailer (Kraiselburd et al., 2004). For example, if the product of one supplier sells out, customers may substitute it for a similar product of a competitors, which may be a strong motivation for suppliers to improve their service levels.
1.3
The state of current empirical research on VMI
In contrast to the extensive stream of research based on analytical models, empir-ical research on VMI is very limited in both quantity and scope. Where modeling studies have attributed differences in VMI performance outcomes to moderating ef-fects related to (aspects of) the replenishment processes, empirical studies on VMI have mainly focused on organizational and relational factors to explain differences in VMI success. Since comprehensive reviews of the body of knowledge on VMI
already exist (e.g. Kauremaa et al., 2009; Marqu`es et al., 2010; Govindan, 2013), we limit ourselves to a general discussion of four strands of empirical work that we identify in the literature.
The first strand consist of the few prior empirical studies that have measured the impact of VMI implementations on supply chain performance using transaction data. Clark and Hammond (1997), Lee et al. (1999) and Dong et al. (2014) all used data from a large supplier that drove the introduction of VMI for its retailers, to study if VMI indeed leads to reduced inventory levels and/or increase service levels at the retailers’ stock points. Because in these settings a large and powerful supplier was driving the VMI implementation, consignment inventory was not included. The outcomes of these studies unanimously show that the supplier was indeed able to lower inventory levels at the stock points of its retailers, with similar or even higher demand fulfillment.
A second strand in the empirical VMI literature focuses on the (perceived) per-formance outcomes and success factors of VMI. This theme has received somewhat more attention in the literature (Kauremaa et al., 2009) and has resulted in the iden-tification of several characteristics of organizations that were successful with VMI, such as, market aspects and power relations (Vergin and Barr, 1999; Claassen et al., 2008) and employee involvement (Kulp, 2002). Many of these studies point out that prior availability of information systems is among the most important factors for per-formance improvement from VMI (Daugherty et al., 1999; Kuk, 2004; Claassen et al., 2008). Others find that it is not the availability of information systems, but the sup-pliers’ competence to work with those systems that affects performance (Vigtil and Dreyer, 2008). In terms of performance, most studies show that organizations that use VMI have higher profit, lower stockouts and cost savings compared to compa-nies that share information alone (Kulp et al., 2004; Claassen et al., 2008). However, other studies, focusing on trust, indicate that VMI is not related to improved supply chain performance (Brinkhoff et al., 2015; ¨Ozer et al., 2017).
A third strand of VMI literature includes four empirical studies investigating antecedents for VMI adoption (Kulp, 2002; Dong et al., 2007; Nagarajan and Ra-jagopalan, 2008; Borade et al., 2012). They identified which context characteristics (e.g., supplier, product and buyer properties) are related to the willingness to adopt VMI. These studies focused on antecedents for VMI adoption (or non-adoption) and are thus not providing insights on the resulting performance outcomes after VMI
adoption. Kulp (2002) emphasizes the retailer’s ability and willingness to share high quality data as an important reason for supplier to adopt VMI. Dong et al. (2007) show that competitiveness of the supplier’s market and buyer–supplier cooperation are positively associated with VMI adoption, while operational uncertainty for the buyer is negatively associated with VMI adoption.
The final strand of empirical studies focuses on different VMI configurations, a theme that has received little attention in the literature. Vigtil (2007) identifies what types of demand data would be valuable for suppliers to improve their planning of replenishment processes in situations with VMI and found that current inventory status and sales forecasts are the most important kinds of information to be made available to the supplier in a VMI relationship. Kauremaa et al. (2009) identify three different patterns that describe the intensity of the VMI implementations in their cases (i.e., basic, cooperative and synchronized), which they relate to supplier and buyer context factors. The patterns suggest that the way in which VMI is used and the performance outcomes that can be expected depend on supplier and buyer con-text. Elvander et al. (2007) propose a taxonomy of possible VMI strategies, formed by implementing different combinations of VMI components. However, how each of those strategies affect replenishment processes, and hence which VMI strategy should be implemented by different suppliers to actually realize improved perfor-mance, remains unknown (Elvander et al., 2007).
When we review these four strands of literature together, we identify three as-pects of VMI implementations that make that there are still gaps in empirical under-standing of how VMI affects suppliers and retailers a supply chain, even though the concept has been around for over 30 years. First, the view on the supply chain per-formance effects of VMI varies across studies, because these perper-formance effects are strongly dependent upon the goals of the VMI introduction in the first place and also on how performance is measured or perceived. Second, the VMI configurations con-sidered in different empirical studies varied widely, and it not always clear which configuration—or transition between configurations—are analyzed, i.e., which com-bination of VMI components is studied. Third, VMI is almost always implemented in an existing context of suppliers and retailers, and it is very likely that this con-text influenced the performance effects of VMI, which complicates generalization of results.
rich the landscape of VMI actually is, but, as a logical consequence, has not been able to link specific configurations to specific performance changes. For one partic-ular VMI configuration, where a large supplier introduces VMI for its buyers, re-sults based on transaction data have provided convincing evidence of how VMI can reduce inventories and improve service levels (Clark and Hammond, 1997; Dong et al., 2014). However, all these studies focus on inventory levels at the stock points of the buyers, leaving changes in inventory levels at the supplier—and inventory in transit—out of scope. This focus clearly matches the goal of these particular VMI in-troductions, yet there are other potential goals to be reached by introducing VMI as well. Furthermore, the context where a single supplier introduces VMI for multiple buyers allows observing how differences in the characteristics of the buyers affects VMI performance outcomes (Clark and Hammond, 1997; Lee et al., 1999), but does not allow for studying supplier characteristics. Explicit performance measures of other VMI configurations (e.g., configurations including consignment inventory or multiple suppliers) are not available in the empirical literature.
Both the strand of literature on the antecedents for the adoption of VMI and that of the (perceived) performance outcomes emphasize the importance of the supply chain context in which VMI is implemented. The contradictions that have arisen across research findings on specific success factors seem related to the (often implicit) way the VMI implementation was expected to improve supply chain performance. While the broad samples of most empirical studies increased external validity of the research findings, and hence the generalizability of those findings, internal validity has been limited and important details on specific VMI configurations are lost. Sim-ilarly, VMI is often the starting point for other (collaborative) efforts in forecasting. The effects of these other forecasting efforts can overshadow the results of the actual decision transfer that is central to VMI (e.g., De Toni and Zamolo, 2005) and make it unclear which transfer between VMI components is exactly observed.
1.4
Research approach and contribution of this thesis
1.4.1
Research objective
The limitations in availability and scope of empirical research outlined in the previ-ous section make that our understanding of how VMI can be used to enhance supply
chain performance is limited. Specifically, empirical understanding of the potential ways in which suppliers can use VMI to improve the performance of a supply chain, i.e., demand fulfillment, inventory and resource utilization, is limited (section 1.2) as is our understanding of how suppliers (with certain characteristics) actually respond to a VMI introduction (Section 1.3). Therefore, the goal of this thesis is to: Empirically study how suppliers can use VMI to improve supply chain performance.
We build on the rich theoretical background based on analytical operational and economic models that outline potential ways in which VMI can enhance the per-formance of a supply chain. In doing so, we take multiple perspectives ranging from a focus on objective observations of the supply chain performance metrics to interpreting the subjective motivations and capabilities of suppliers. In this way, we contribute to research streams that employ inventory measure in order to study sup-ply chain performance, but also more qualitative focused research streams that look further into the reasons behind these observations.
1.4.2
Research setting
Our research goal requires an approach and setting that is unique in relation to prior work on VMI in two aspects in particular. First, we aim to capture all relevant as-pects of why and how organizations adapt their replenishment process in response to a VMI introduction, and how these adaptations affect their performance over time. This requires collecting and studying a combination of rich qualitative and quanti-tative data on both the supplier and retailer side. To link the different VMI com-ponents to observed supply chain performance outcomes, this thesis aims to obtain in-depth understanding of a single research setting, rather than generic insights from a broader sample. Second, to understand how a supplier’s context affects their re-sponse to VMI, we need to study a research setting in which suppliers operate under the same VMI conditions. Studying suppliers that deliver the same retailer under the same contracts and with access to the same information allows for a ’fair’ compar-ison between suppliers. A limitation of such a research setting is that only a single VMI configuration and retailer context is observed. However, we aim to use our in-depth understanding of the research setting to obtain insights into how the products and organizational characteristics of different suppliers affect their VMI adoption decisions and performance outcomes, which can be generalized to other settings.
In this way, our study falls in the realm of previous studies that have used real-life data from a single research site (Clark and Hammond, 1997; Dong et al., 2014; van Donselaar et al., 2010).
The empirical data for this thesis has been gathered by observing an introduction of VMI with consignment inventory by a large European retailer. The retailer intro-duced VMI for all of its 100+ fresh food suppliers. Before, after and during this tran-sition, the retailer provided us with access to uniquely rich data. We were able to be embedded in the context from the beginning of the VMI introduction to 3 years fol-lowing this introduction for the sole purpose of this research. During this period, we conducted interviews at multiple points in time with employees of the retailer and 10 of the suppliers and, based on these interviews, collected additional survey informa-tion on the full sample of suppliers. In addiinforma-tion, we collected extensive transacinforma-tion data describing the inventory and service level performance of the entire fresh food supply chain. The availability of this rich data enabled a multi-method, longitudinal approach for the thesis. All chapters make use of both quantitative and qualitative data, but the extent to which differs based on the scope and methods applied.
1.4.3
Research approach
The studies presented in this thesis all apply empirical methods to study how sup-pliers can adopt VMI and how the context they operate in affects their decisions and abilities to do so. The first chapter focuses on general performance effects around the transition to VMI, while the following studies zoom in on how suppliers use VMI in the long term and how this is conditional on their characteristics. This way, the find-ings of the studies complement each other to provide a larger, more complete, view of how suppliers can use VMI.
Chapter 2 presents a study of how a VMI introduction driven by a retailer affects the performance of the supply chain as a whole before, during and after a transition to VMI. The study shows how VMI introduced by a large retailer is different from VMI introductions by large suppliers, as seen in the literature thus far.
Motivated by the findings of Chapter 2, Chapter 3 presents a study in we test different mechanism through which VMI can improve supply chain performance using transaction data. The results show that suppliers can use VMI to improve the performance of a supply chain in two different ways, depending on their product’s
characteristics.
Chapter 4 presents a longitudinal case study of 8 suppliers-retailer dyads in which we study how VMI introduction affect suppliers’ and retailers’ organizations and the relation between them. We zoom in on how the suppliers’ organizational motiva-tions and abilities develop over time, and show how this affects their service levels and relation with the retailer over time.
Finally, Chapter 5 concludes and summaries the main research findings, dis-cusses their limitations and generalizability, and provides an outlook for future re-search on the topic.
Chapters 2-4 of this thesis are based on papers that have been submitted to peer-reviewed journals or are in the process of being submitted. The paper-based struc-ture has resulted in some repetition from chapter through chapter to make them readable as separate entities.
• Chapter 2: Post, R.M., Buijs, P., Wieringa, J.E., & Wortmann J.C. ‘The per-formance effects of retailer driven Vendor-Managed Inventory with Consign-ment’
• Chapter 3: Post, R.M., Buijs, P., Wieringa, J.E., Fransoo, J.C, & Wortmann J.C. ‘The two ways in which Vendor-Managed Inventory can improve supply chain performance’
• Chapter 4: Post, R.M., Buijs, P., & Wortmann J.C. ‘A cooperation and coordina-tion perspective on supply chain collaboracoordina-tion dynamics’
Next to this, the research resulted in three conference papers and two short arti-cles in professional literature.
• Post, R.M., Buijs, P., & Wortmann J.C. 2016. The effect of vendor context on Vendor-Managed Inventory benefits: A conceptual framework. 23th EurOMA Conference proceedings
• Post, R.M., Buijs, P., Wieringa, J.E., & Wortmann J.C. 2017. Under what con-ditions should the supplier own and decide on inventory for the best supply chain performance: an empirical study. 24th EurOMA Conference proceedings • Post, R.M., Buijs, P., & Wortmann J.C. 2019. A Coordination and Cooperation
perspective on Supply Chain Collaboration Dynamics. 80th Annual Academy of Management conference proceedings
• Post, R.M. 2017. Operatie op de magazijnvloer belangrijk voor succes VMI. LogistiekProfs
