• T.C. Edwin ChengEmail author
  • Jian Li
  • C.L. Johnny Wan
  • Shouyang Wang
Part of the International Series in Operations Research & Management Science book series (ISOR, volume 143)


Postponement has been discussed by numerous scholars and researchers from different perspectives since 1950s when Alderson [3] observed that products tend to differentiate when they near the time of purchase in order to reduce marketing costs. He named this concept the principle of postponement. Postponement , also known as late customization or delayed product differentiation, refers to delaying some product differentiation processes in a supply chain as late as possible until the supply chain is cost effective (Garg and Lee [43]). It gives rise to economies of scale and scope through product and process standardization and customization, respectively.


Supply Chain Product Family Mass Customization Customer Order Economic Order Quantity 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Postponement has been discussed by numerous scholars and researchers from different perspectives since 1950s when Alderson [3] observed that products tend to differentiate when they near the time of purchase in order to reduce marketing costs. He named this concept the principle of postponement. Postponement , also known as late customization or delayed product differentiation, refers to delaying some product differentiation processes in a supply chain as late as possible until the supply chain is cost effective (Garg and Lee [43]). It gives rise to economies of scale and scope through product and process standardization and customization, respectively. The goal of postponement is to supply desirable products to customers at a relatively low cost and in a responsive way (Feitzinger and Lee [41]). Li et al. [73] selected postponement as one of five major SCM practices that have discernible impact on competitive advantage and organizational performance. In fact, postponement is an element of mass customization that is applied to cope with product variety in order to enhance customer quality. A comprehensive literature review of how service quality is improved by an adoption of postponement to deal with a wide product range is presented in this chapter.

This chapter is organized as follows. In Section 1.1 the relationships between product variety , mass customization and postponement are discussed. In Section 1.2 four types of postponement are addressed. The advantages and disadvantages of postponement strategies and the prerequisites for postponement strategy development are also discussed. In Section 1.3 a review of the cost models for analyzing various postponement strategies is provided. A literature review for our model development is presented in Section 1.4. In Section 1.5 we conclude the chapter.

1.1 From Product Variety to Postponement

1.1.1 Product Variety

MacDuffie et al. [75] defined product variety as the breath and depth of product lines. Strategically, broadening product variety gives a company a distinctive competitive advantage in quickly responding to ever-changing market environments and customer tastes. Product proliferation can lead to greater flexibility, larger market shares, deeper market penetration, higher customer satisfaction and loyalty. However, it increases the number of set-up and inventory level, and demands more technical services, material handling, supervision, co-ordination and support in production (Yeh and Chu [128]). Since expanding product variety increases both values and costs, only value-added variety should be offered.

Product variety falls into two dimensions: perceived variety and actual variety. Customer value is only enhanced by an increase in perceived variety, regardless of the extent of actual variety offered (Porter [94] and Kahn [60]). Adding actual variety (but not perceived variety) not only increases customer confusion but also raises costs. Stalk [109] identified two forms of costs: those responding to volume or scale, and those driven by variety. The former reduces when volume increases, while the latter increases when manufacturing becomes complex as variety increases.

Since handling product variety is a complicated process, Ulrich et al. [115] identified 12 supply chain decision areas and classified them into strategic level and tactical level for coping with product variety . Strategic decisions include: (1) dimensions of variety offered, (2) distribution channel, (3) degree of vertical integration, (4) process technology, (5) position of decoupling point, and (6) product architecture. Tactical decisions cover: (1) number and combination of product attributes, (2) extent of parts sharing, (3) lot size policy, (4) inventory management policy, (5) production scheduling, and (6) promotion plans. Besides, information about product lines, customer behavior and tastes, market segmentation, suppliers, technological innovation, and strengths and weaknesses of competitors is vital for product variety strategy planning (Porter [94]).

1.1.2 Mass Customization

As mentioned above, an effective product variety strategy calls for the provision of highly perceived variety of products and services to individual customers. From a supply chain management’s point of view, it aims at supplying customized products and services to each customer in a responsive and cost effective way. It is important to note that customization is a time consuming and costly strategy due to diseconomies of scale and scope. The associated costs include the logistics of managing variety, material handling, quality, production capacity and inventory (Martin et al. [76]). Mass customization (Feitzinger and Lee [41]) or customized standardization (Lampel and Mintzberg [63]), are aimed at balancing both standardization and customization in order to achieve both quick response and low cost. It is a guiding strategy for those multinational companies that manufacture global products with customization for local markets.

Lee [65], Feitzinger and Lee [41], and Yeh and Chu [128] mentioned three prerequisites for mass customization. The first is modular design of production processes. Processes can be moved, assembled and re-arranged to support various products manufacturing easily and at low costs. Moreover, modular design allows concurrent production, isolates potential problems within module, and most importantly, enables postponement. One of these process-re-engineering techniques is operations reversal (Lee and Tang [70]). It is an approach in which the sequence of two consecutive processes in a supply chain is reversed to reduce demand variability. The second requirement is products and parts standardization . Homogenous products are produced and they are capable of supporting multiple product functions and features with slight re-configurations. It reduces costs in part number administration, leads to inventory reduction and facilitates supplier management (Lee [64]). The last building block is a flexible supply chain , which involves the coordination and negotiation of marketing, R&D, manufacturing, distribution, finance and retailing, to support both generic and customized products responsively (Lee [64]).

1.1.3 Postponement Strategy

Garg and Lee [43] suggested applying a postponement strategy to deal with product variety . The concept was first discussed by Alderson [3]. In his principle of postponement , he argued that products differentiation in the point of purchase could reduce various marketing costs. In this regard, he suggested all changes in form and identity be delayed to the latest possible point in time and location. Christopher [29] referred to postponement as a vital element in an agile strategy, which adopts both a flexible manufacturing system (FMS) and standardization of products to achieve organization flexibility.

1.2 Classification of Postponement

Postponement , also known as late customization or delayed product differentiation, refers to delaying some product differentiation processes in a supply chain as late as possible until the supply chain is cost effective (Garg and Lee [43]). It implies economies of scope and scale can be achieved by product and process standardization . Economies of scale are made possible through standardization of components and processes to support a large variety of products. Economies of scope are achieved by producing various products at the same time (Pine [93]).

Under a postponement strategy, products from a product family share common parts and processes until their point of product differentiation. After the point of product differentiation , a “fan-out” occurs, as end products require different components and processes (Garg and Lee [43]). This strategy benefits from the “risk-pooling” effect, which suggests demand variability is reduced by considering aggregate demand instead of individual demand in a product category (Federgruen and Zipkin [40], Simchi-Levi et al. [105]). The major reason is that one extremely high demand may be offset by another extremely low demand after aggregation. Lower demand variability implies fewer safety stocks, lower inventory levels and more accurate resources planning. Apart from reducing demand variability, postponement is also used to tackle process and supply uncertainties.

There are four forms of postponement strategies, namely pull postponement , logistics postponement , form postponement (Lee [65]) and price postponement (van Mieghem and Dada [124]). The former three strategies are also referred to as production postponement (van Mieghem and Dada [124]).

1.2.1 Pull Postponement

Companies usually employ forecasting techniques to estimate customer demand. Products are produced in advance of customer orders, where production is planned to achieve optimal capacity and efficiency. Such a production strategy is called a make-to-stock (demand-push) strategy . One advantage of this strategy is that it has immediate stock availability (Schroeder [101], Browne et al. [19] and Arnold [5]) to respond to customer orders quickly. However, unavoidable overstock or stock outs occur due to demand forecast variations. The variation is amplified from downstream to upstream in a supply chain due to information distortion. This phenomenon is described as the “bullwhip effect” (Lee et al. [68]). Stock outs lead to losing customers while overstocked items become obsolescence at the end of the product life cycle. The costs associated with overstock and stock outs can be huge. In order to cope with these unwanted variations, a make-to-order (demand-pull) strategy is advocated.

In contrast to a make-to-stock strategy , a make-to-order strategy pulls production once a customer order is received. It can entirely eliminate unwanted inventory as production quantity is set after demand uncertainty is resolved (van Mieghem and Dada [124]). However, long order lead-time and wide production fluctuations are two major drawbacks of this strategy.

In practice, most companies operate between these two extremes in order to balance production capacity and demand (De Hann et al. [35]). A customer order decoupling point (Hoekstra and Romme [55], Browne et al. [19]) or a push-pull boundary (Brown et al. [17]), which indicates the extent of which a customer order penetrates into the goods flow, separates forecast-driven activities upstream from demand-driven activities downstream in a supply chain system (van Donk [117], van Hoek [118]). If a system cannot be decoupled, it should be operated only in either a fully forecast-driven or fully demand-driven mode (Hoekstra and Romme [55]). The concept of a customer order decoupling point was proposed by Bucklin [21] in 1965. Bucklin developed a postponement-speculation approach such that inventory is only built at each process in a distribution channel whenever the costs are less than the savings among channel members.

Pull postponement , also known as process postponement (Brown et al. [17]), refers to moving the decoupling point earlier in the supply chain such that fewer steps will be performed under forecast results (Lee [65]). There are two decoupling points: one for the physical supply chain and the other for the virtual supply chain (Christopher [29]). The virtual supply chain refers to the information-sharing network among supply chain partners. In pull postponement , customer demand is the key information for determining the push–pull boundary of a physical supply chain . In fact, it is an information strategy (Lee [65]).

For instance, assume there are four processes in a supply chain where process 4 receives customer order (see Fig. 1.1). If process 2 and process 3 update the customer order information simultaneously, the decoupling point for the virtual supply chain is in process 2. However, the pull postponement strategy can be carried out in either process 2, process 3 or process 4 but not in process 1 because it cannot receive customer order information for pull production. Process 1 can employ a push system only. In theory, the best pull postponement is implemented in process 2 as it is the earliest process that can be operated once an order is received. However, if the process lead-time for process 2 through process 4, L, is longer than the customer expected waiting time W, the decoupling point should be moved downstream to process 3 or 4. Otherwise, customers need to wait extra time for the product or they take their businesses to other suppliers.
Fig. 1.1

A four-step supply chain process

Customer Order Decoupling Point and Point of Product Differentiation

Recall that a postponement strategy aims at moving the point of product differentiation as late as possible until it is cost effective. On the other hand, the purpose of a pull postponement strategy is to move the decoupling point earlier to minimize forecasting error. Thus, they seem to be unrelated and contradicting concepts. However, they are in fact related and vital for a postponement strategy to be carried out successfully. There is little literature discussing their relationships.

Rethink the four-process supply chain again in Fig. 1.1. This time the supply chain is analyzed from the product, instead of the process, perspective. Assume that there are two products, A and B, whose distinctive features are added after the completion of process 2. The revised diagram is shown in Fig. 1.2.
Fig. 1.2

A revised four-step supply chain process

Assume further that information is shared and updated throughout the supply chain . If the decoupling point D lies in process 1, the customer waiting time would be L 1. It is a pure make-to-order system. In practice, production volume fluctuates widely due to demand variability and all orders are backlogged. In order to smooth the production schedule and make the system more responsive, D is moved to process 2 such that process 1 is make-to-stock . The customer waiting time is \(L_{2} <L_{1}\). Since inventory has been built up after process 1, there is little intention to stop passing process 2 in a series supply chain system until customer orders are resolved. In this sense, D will be further passed to process 3 and the customer waiting time is shortened to L 3. If D is further moved to process 4, demand variability increases as products start to differentiate after process 3. There are risks of overstock and stock outs of products A and B if D is set after the point of differentiation K. As a consequence, D should be set before K in order to implement postponement. D is determined by customer expected waiting time and availability of demand information in the system, while K is determined by operational factors such as manufacturability, product characteristics, costs and so on.

Applications of Pull Postponement

National Bicycle differentiates themselves from other traditional bicycle manufacturers by adopting a different decoupling point. They move the decoupling point from final assembly to frame welding such that they could offer more than 11 million options to customers quickly (Lee [65]). Xilinx employs a pull postponement strategy in semiconductor manufacturing (Brown et al. [17]). Dies are transformed to specialized integrated circuits (ICs) once customer orders are received. Xilinx management needs to monitor 100 types of work-in-process (WIP) inventory of dies, instead of 10,000 finished products. Corporate inventory drops by 23% after implementation for 1 year (Brown et al. [17]). Besides, Wings & Legs, a large poultry processor in the Netherlands, delays their packaging and labeling processes as their key clients demand tailor-made packages and labels for Wings & Legs’ chicken legs (van Dijk et al. [116]). Benetton, an apparel manufacturer, postpones its color dyeing process until orders are received (Lee and Tang [70]). They all enjoy substantial savings through inventory reduction.

1.2.2 Logistics Postponement

Logistics postponement involves the re-designing of some of the processes in the supply chain so that some customization can be performed downstream closer to customers (Lee [65]). Bowersox and Closs [14] defined logistics postponement as an emerging strategy of form , time and place postponement . Similar to pull postponement, their meanings of form and time postponements refer to delaying some customization activities until customer orders are received. Place postponement refers to the positioning of inventories upstream or downstream. In general, logistics postponement is an extension of pull postponement .

Logistics postponement considers whether pull postponement can be implemented more effectively and efficiently by relocating some demand-pull processes closer to customer levels (Lee et al. [67] and Lee [65]). It is associated with the concepts of design for localization or design for customization, which takes operational and logistics services into account in the design process so as to serve different market segments (Lee et al. [67]). Packaging postponement and labeling postponement (Twede et al. [114]) or branding postponement (Ackerman [2]) can be subsets of logistics postponement when the packaging, labeling or branding processes are moved closer to customers.

Applications of Logistics Postponement

Hewlett-Packard produces generic printers at its factory and distributes them to local distribution centers, where power plugs with appropriate voltage and user manuals in the right language are packed. Since generic printers are lighter, more units could be shipped. Distribution cost is cut by half, and million dollars have been saved, although manufacturing costs are slightly higher due to the use of standardized components to support mass customization (Lee [65], Lee et al. [67], Feitzinger and Lee [41]).

WN, a European wine producer, demonstrates another real-life example of logistics postponement application. WN produces base wine in central bodega and defers bottling, packaging and labeling activities at the local level (van Hoek [118]). Ackerman [2] quotes a similar strategy adopted by Coca-Cola, in which concentrated syrup is shipped to retailers where it is mixed with carbonated water to form Coca-Cola in retailers’ soda fountains.

Twede et al. [114] presented a logistics postponement application at Swedish furniture retailer IKEA. All products in IKEA retail stores are kept in semi-finished forms (flat packs) and are assembled by customers or deliverymen after home delivery. In this way, truckload capacities can be utilized and configurations can be easily made at customer locations. Kellogg Company, the world’s leading cereal food producer, ships its products in bulk to local co-packers, where some constituents are added and packed (Brown et al. [18]).

1.2.3 Form Postponement

Form postponement, also called product postponement (Brown et al. [17]), opts for a fundamental change of the product structure by using standardized components and processes to achieve high customization (Lee [65] and Brown et al. [17]). Bowersox and Closs [14] introduced form postponement as a postponement of the final manufacturing or processing activities. Their concept is akin to Lee’s [65] definition of pull postponement. In order not to mix up the two concepts, the version of Lee [65] and Brown et al. [17] is adopted. In fact, form postponement is an enabling strategy to supplement pull postponement (Lee [65]).

Applications of Form Postponement

Brown et al. [17] applied form postponement in a semiconductor company (Xilinx), where it re-designs the IC so that it could be re-configured by software easily and quickly for customized features and functions. It is particularly useful in programmable devices because a nearly infinite number of products can be produced by using program configuration.

1.2.4 Price Postponement

Van Mieghem and Dada [124] defined price postponement from economic and marketing perspectives. They described price postponement as a strategy aimed at deferring the pricing decision until customer demand is known. Selling price is negotiated with customers after they place their orders. Based on their findings, one advantage of the price postponement strategy is that it makes investment and production decisions insensitive to demand uncertainty, since profit margin can be covered by setting various selling prices after demand is known. Another advantage is its ease of implementation. Unlike the above three postponement that require re-engineering techniques such as operations reversal and standardization of product and process, price postponement is a managerial decision that is determined by marketers.

Applications of Price Postponement

Bank of China (BOC) Hong Kong applied a price postponement strategy for its initial public offering in Hong Kong in July 2002. In the face of high demand uncertainty, BOC set an offering price range, between HK$6.93 and HK$9.5 per share, for investors to subscribe to its shares that were worth HK$6.93 per share (South China Morning Post, 23 July 2002 [108]). Since the public offering was over-subscribed by 26 times, BOC finally allotted at least 500 shares to each investor at a price of HK$8.5 (South China Morning Post, 22 July 2002 [107]).

1.2.5 Implications

Practically, postponement strategies can be combined, integrated or partly applied to a supply chain in order to achieve different objectives. Recall the model shown in Fig. 1.2, if the aggregate demand of products A and B is known and standardized components are ready in process 2 (point of differentiation) for customization in process 3 and process 4, both form postponement and pull postponement strategies are applied. Furthermore, if process 3 and process 4 are performed locally for products A and B, instead of in central production facilities, logistics postponement is adopted. In selling, if product prices are set after customer orders are received for product A and product B respectively, a price postponement is employed. This example is flexible to be applied in all supply chain systems.

Not all products can be made shortly after customer demands are known. If they are not available within a promising time, customers may turn down the business. Therefore, there is a need to keep some inventory for these products, while a make-to-order approach is used to those products that can be produced within customer expected waiting time. This mixed strategy is referred to as a partial postponement strategy or a hybrid postponement strategy (Brown et al. [17], Graman and Magazine [50]), in which the lead-times after the point of differentiation and customer expected waiting time are key determinants. Xilinx keeps both semi-finished dies and finished IC inventory in dealing with long back-end lead-time of some products (Brown et al. [17]).

1.2.6 Advantages and Disadvantages of Postponement

Advantages of Postponement

To some extent, the philosophy of postponement follows the JIT principle , as both emphasize to have the right product in the right place at the right time (Cheng and Podolsky [27], and Heskett [52]). In fact, postponement offers substantial advantages for a supply chain to improve in terms of time, quality and cost. Graman and Magazine’s study [50] found that although postponement results a reduction in inventory in terms of quantity, the service level is unchanged. In general, less inventory held makes inventory management easier and more responsive. On the other hand, perceived product quality is enhanced by small design changes (Lee [64]). Besides, standardized components can reduce the risk of obsolescence since configurations become easier in the form of WIP inventories, instead of end products (Brown et al. [17]).

Postponement makes forecasting easier at a generic level than at the level of finished forms because demand variability is reduced by aggregation (Christopher [29], and Ernst and Kamrad [38]). It is particularly obvious under a multi-echelon supply chain system in which the demand of the current stage is equal to the demands of the previous stages. Moreover, it supports various production alternatives such as engineering to order, purchasing to order, make to order, manufacture/assemble to order, packaging and labeling to order, shipment to order and adjust to order by shifting the customer order decoupling point (van Hoek [118], Olhager [89], and Hoekstra and Romme [54]). Its modularity characteristic not only reduces the cost of assembly (Chiou et al. [28]) but also enables outsource capability and speeds up new product development (Brown et al. [17], and Ernst and Kamrad [38]). The higher degree of modularity, the more outsourcing opportunities a company can pursue. Thus, fixed investment can be reduced drastically. In particular, logistics postponement can initiate the use of third party logistics (3PL) to handle local value-added activities and product delivery (van Hoek [120]). Successful outsourcing examples are Dell Computer, Nike, Reebok and General Motors (Tully [113]).

Disadvantages of Postponement

Notwithstanding postponement brings significant benefits to a supply chain, the cost of re-engineering and developing the supply chain cannot be neglected. Since a postponement strategy aims at delaying customized activities as late as possible until customer orders are received, more work-in-process inventories need to be built up before the point of differentiation (Brown et al. [17]). To a certain extent, standardized components increase variable costs because they need to support various product features (Lee [65], Ma et al. [74]). Power plugs, for example, should be re-designed so that a switch is added in order to support different voltage supplies in different countries. The modification makes variable costs higher. However, variable costs can be reduced by shifting the assemble processes to local facilities so that there is always one suitable type of power plug available for the product. In this scenario, other factors such as transportation cost, setup cost, training cost and local material cost should be weighted against variable cost savings.

Moreover, there is always a trade-off between mass production and customization in carrying out postponement as the former gains economies of scale while the later gains higher customer values. Economies of scale are lost after the point of differentiation due to customization (Zinn and Bowersox [130]). This effect is more pronounced in logistics postponement since customized processes are performed separately in local facilities with different product lines. To a large extent, it is associated with a high risk of creating quality problems as production is moved from central facilities to local facilities or even at the retailer level (Ackerman [2]).

1.2.7 Prerequisites for Postponement Strategy Development

Postponement is not a panacea for all industries, as it may not be possible or economical for companies to re-design common processes and components such that specific end products are produced from a group of generic products. Recall that Lee [65], Feitzinger and Lee [41], and Yeh and Chu [128] mentioned there are three building blocks for carrying out mass customization . Lee [65] described four postponement enablers, namely modularity, design for postponement, supply chain collaboration and associated costs. Besides, there are additional prerequisites that are conducive to postponement strategy development. All prerequisites are summarized below:
  1. (i)

    Mass customization principles should be embedded throughout the supply chain system such that the three building blocks are present to support and facilitate postponement implementation.

  2. (ii)

    Products can be categorized into various product families such that each product family shares common characteristics in terms of product design, standardized parts, common production processes and the same production location. In other words, a point of product differentiation can be defined. Bills of materials (BOM) help one to look for opportunities to group common parts at a lower level of the product structure so that products in the same product family share more common production processes (Kennedy et al. [62]).

  3. (iii)

    Operations should be closely linked with product and process design (Ma et al. [74]). That is, the philosophy of design for flexible manufacturing is essential for effective postponement strategy development. The concept of postponement should be embedded in the design process such that not only cost, quality, flexibility and serviceability but also distribution, service, maintenance, marketing, manufacturing capabilities, inventory management and supplier management are considered (Barkan [11] and Calvin and Miller [23]).

  4. (iv)

    Postponement is particularly powerful if the supply chain network operates on a global scale in which the positioning of inventory, production mode and structure, and distribution facilities become critical success factors for both cost reductions and customer value creation. Global efficiency can be achieved through mass production while local responsiveness can be enhanced by customization (van Hoek et al. [123]).

  5. (v)

    Lead-times after the point of postponement should be justified with respect to customer expected waiting time for the product to avoid backorders and lost sales (Lee [64]). Under certain circumstances, partial postponement may be used to reduce lead-times.

  6. (vi)

    Information technology and greater supplier involvement are vital factors to streamline the supply chain process (Brown et al. [17]).


1.3 Cost Models for Analyzing Postponement Strategies

In general, models for analyzing postponement strategies can be classified into four types. They are deterministic models , stochastic models , heuristic models and descriptive models based on case studies (Beamon [12]).

1.3.1 Stochastic Models

We define a stochastic model as an inventory model where demand in any period is random (Hillier and Lieberman [53]). Lee and Tang [69] formulated a total relevant cost model to analyze the effectiveness of a designed strategy to be applied in a N-stage manufacturing system that produces two products, whose demands follow normal distribution . The first k operations are common for the two products. Their cost model consists of four cost factors, including total average investment cost, total processing cost, total WIP inventory cost and total buffer inventory cost. As an extension, Garg and Tang [44] considered a production system that has two product differentiation points. Their cost model showed that these two points also yield a lower inventory saving. Other cost models for analyzing the point of product differentiation can be found in Garg and Lee [43] and Lee [64].

Ernst and Kamrad [38] developed a total cost model to analyze four supply chain structures, namely rigid, modularized, postponed and flexible, of an ice-cream supply chain that serves two different markets. Demand of a particular flavour follows a probability density function, and the total cost under evaluation includes fixed cost, variable cost, holding cost and backorder cost. Surprisingly, they concluded that postponement is the worst choice among four. However, this conclusion is drawn directly from numerical analysis of an ice-cream supply chain. A more generalized model should be developed in order to provide a more concrete and reliable framework for evaluating supply chain effectiveness under various scenarios.

Aviv and Federgruen [6] modeled a two-phase production system in which common products are produced in the first phase and product differentiation is delayed to the second phase. Their objective is to minimize an expected long-run average discounted cost of j products, each of them follows multivariate distribution with arbitrary correlations. First, they found a lower-bound of the average cost for a single-stage single-product production. Then they extended it to a two-stage postponement system with a heuristic strategy. A numerical analysis and a study of Hewlett-Packard case are provided. They found that postponement results in substantial savings when coping with a large degree of product variety , and less correlated or high seasonality product demands.

Ma et al. [74] analyzed component commonality and postponement in a multi-stage multi-product assembly system. A set of common base-stock levels at all stocking points is found to minimize the total expected inventory cost. In their model, both lead-time and service level are considered. They showed that those processes with long procurement lead-times are rearranged after the point of product differentiation so as to meet the service level . Also, component commonality is more preferred to be implemented in earlier stages. In fact, the relationships between part commonality and aggregate safety stock are discussed by Collier [32], McClain et al. [78] and Baker et al. [9]. They all supported the view that less aggregate safety stock is needed for maintaining a constant service level .

1.3.2 Heuristic Models

Heuristic models are models that employ rule of thumb to approach the best solution (Ballou [10]). Brown et al. [18] presented an application of an enterprise resources planning (ERP) system in Kellogg Company. The system is called Kellogg Planning System (KPS). It is a rule-based heuristic program, which helps Kellogg Company to control its operations, production, inventory and distribution for breakfast cereal and other food products. The objective function is to minimize production costs, packing costs, inventory costs, shipping costs and penalty costs for overstocks and under stocks. Corry and Kozan [33] developed a push/pull hybrid production system (HIHPS) of a foundry from which a single assembly stage is demand-pulled. Simulated annealing algorithm is applied to determine the optimal buffer reorder point and replenishment level of the components that can minimize a total cost function. Another HIHPS system can be found in Cochran and Kim [31]. Graman and Magazine [50] studied a manufacturing system that keeps both WIP and finished goods of a group of products so as to fulfill a given service level . They used a Monte Carlo integration method to determine the stock levels of both inventories. Besides those dynamic systems, Zinn [129] used a percent saving ratio to evaluate the safety stock savings resulted from postponement. He showed that a large product line yields a high safety stock saving when the demand and standard deviation of demand for each product are independent and approximately equal respectively. A simulation study by Johnson and Anderson also revealed that postponement improves fill rate and service level when demand for each product is at the same level [59].

1.3.3 Descriptive Models

Bucklin [21] proposed a postponement-speculation model to test six hypotheses that associated delivery time, product type, product cost and demand variability with the choice of postponement. He claimed that the point of postponement-speculation appears in a distribution channel whenever there are systemwide net savings resulted from postponement. It is one of the earliest research studies on postponement.

Van Hoek [118] made use of a case study approach to compare two logistics postponement strategies implemented in European wine producer WN in terms of transport cost, production cost, material cost, inventory holding cost and bottling cost. The first alternative is to delay final distribution to customer and the second one is to defer the bottling, packaging and labeling activities to the local level. He found that there are cost savings in transportation and inventory holding for both postponement alternatives. In addition, he pointed out that product characteristics, such as product value, volume and weight, affect the choice of postponement. In other papers [120, 121], he studied the role of third party logistics providers (3PL) in carrying out parts of the postponed supply chain activities. He anticipated that 3PL would play a crucial role in those postponed activities that in turn accelerate the adoption of postponement.

Zinn and Bowersox [130] conducted a discriminant analysis to explore factors affecting the choice of five postponement strategies, namely labeling, packaging, assembly, manufacturing and time postponements . They found that cost saving, product value and demand uncertainty are key drivers of postponement. Chiou et al. [28] employed factor analysis and path analysis to explore factors affecting postponement and the causal relationships between demand characteristics and postponement among Taiwanese IT firms.1 They found that modular product design, component cost and product life cycle induce different choices of postponement. Also, Huang and Lo [56] described a postponed PC supply chain in Taiwan that sourced components globally and assembled locally. Jahre [57] applied postponement in a household waste collection process, which provides insights into postponement being applied in the context of reverse logistics to deal with environmental issues. In his model, the waste-sorting process (separation of tin cans, plastic bottles and paper for recycling) can be either at the consumer level or delayed until the waste reaches the collection center.

1.3.4 Performance Measures

In summary, a set of performance measures need to be adopted to evaluate any analytical model formulated to study a particular postponement strategy. Basically, performance measures can be grouped into three categories, namely qualitative, quantitative and time measurements (Beamon [12]).
  • Qualitative factors
    • Service level (Ma et al. [74], Graman and Magazine [50]), product quality (Martin et al. [76]) and information and material flow integration (Martin et al. [76] and Nicoll [85])

  • Quantitative factors
    • Cost minimization
      • Fixed costs, variable costs, inventory costs, distribution costs, per-unit penalty for stock out and overstock (Bucklin [21], Ma et al. [74], Aviv and Federgruen [6], Ernst and Kamrad [38], Lee and Tang [69])

    • Value maximization
      • Sales, profits (Beamon [12])

  • Time
    • Fill rate (Graman and Magazine [50]), order and production lead-time (Ma et al. [74]) and customer order response time (Lee [64])

1.4 A Literature Review for Model Development

To the best of our knowledge, there is only a small number of postponement models that are based on deterministic demand. Recent deterministic models include, among others, those by Wan [125], and Li et al. [71, 72]. There are new research opportunities in studying postponement by deterministic models such as economic order quantity (EOQ) and economic production quantity (EPQ) models. Both models can be used to derive a total cost function for analyzing postponement. They can help to analyze whether or not combining certain supply chain processes can reduce the total cost. One of the potential savings is from joint ordering [119] or joint production of a group of products whose demands are deterministic. EOQ and EPQ models, which relate to joint ordering and joint production, are reviewed and they provide many useful ideas and insights for our model development in the sequel.

1.4.1 EOQ and EPQ Models

The EOQ model is used to answer two questions: when and how many to order (Zipkin [131] and Hadley and Whitin [51]). Silver [104] developed a simple cycle policy based on the EOQ model to decide when to jointly replenish N groups of products, in which one group is replenished every T years while the other N – 1 groups is reordered every \(k_{n}T\) years, where k n is the number of integer multiples of T for the replenishment of item n. The idea originated from Shu [103] and Nocturne’s revised version of Shu’s optimal ordering frequency (Nocturne [86]). Their models only considered two groups of products. In addition to dealing with the issue of when to order, equal attention should be paid to the other issue: how many to order? An integrated EOQ model was presented by Cheng [26], and Chen and Min [25]. They considered profit maximization when multi-products are ordered jointly, subject to storage space and inventory investment constraints. They employed the Karush-Kuhn-Tucker (KKT) conditions to solve the problem. They assumed that there are no backorders .

On the other hand, there are numerous articles addressing multi-product production with EPQ. However, none of them focuses on comparing the total average cost between a postponement system and a non-postponement system . EPQ models for multi-products can be classified into two types: single machine and multi-machines. The single machine EPQ model follows a rotation cycle policy, by which end-products are produced in sequence in each production cycle [83]. Eilon [37] classified the production of several products by a single machine, in which product demand is known and only one product is produced at a time, as a multi-product batch scheduling problem. Instead of grouping the production, he split the batch into subbatches and compared the total cost per day. Goyal [46] studied a similar problem. He used a search procedure to determine the EPQs of two items. The classical problem is found in Tersine [112] and Nahmias [83]. Recall the logic behind a postponement strategy is to group certain processes together so as to lower the total cost. It is more appropriate to consider grouping the production of end-products instead of splitting their production across a production cycle.

Apart from the single machine scheduling problem, the multi-machine multi-item EPQ models fall into the other end of the spectrum. In dealing with this kind of problem, it is not uncommon to group a large number of end-products into different product families based on processing similarities and economic considerations so that each product family shares a single lot size [22]. It can greatly reduce manufacturing complexity. Byrne [22], and O’Grady and Byrne [87, 88] employed a simulation approach to find good production lot sizes for a number of product families that share common machines in order to minimize a total cost function. In their model, each product family is assumed to have a single lot size. A more dynamic model was presented by Bertrand [13]. He developed a total cost model for a production system that produces multi-items in multi-work centers. His model is based on batch size optimization and queuing theory and it accounts for both finished goods and WIP inventories. Good solutions are found by using the Newton-Raphson method. Besides, Goyal et al. [49] presented a realistic problem in determining the EPQ at each production stage for multiple items across a multi-stage production system. However, due to problem complexity, their model can only be solved by a heuristic approach that yields sub-optimal solutions.

1.4.2 Lot Size-Reorder Point Model

An (r,q) inventory policy , also known as a lot size-reorder point model (Nahmias [83], and Hadley and Whitin [51]), attempts to optimize a total cost function by continuously reviewing the inventory level in order to fulfill stochastic demands. In this system, r and q are independent decision variables. The operation of an (r,q) inventory policy is that: when the inventory position drops to a reorder point r, an order of fixed quantity q is placed (Federgruen and Zheng [39]). Nahmias [83] constructed an expected average annual cost function that includes fixed setup cost, inventory holding cost and shortage cost. His cost function used estimated average inventory as he found that the true average inventory is complicated to derive. He assumed demand is normally distributed and an iterative procedure is employed to solve for optimal values of r and q, starting with \(q=\mathrm{EOQ}\). It is similar to the model developed by Hadley and Whitin [51]. Moreover, Hadley and Whitin [51] developed an exact cost function for solving a case in which demand follows a Poisson distribution , lead-time is constant and all unfilled demands are backordered. A searching method is required for solving their model. Generally speaking, there is no ‘reliable and straightforward method’ for solving an optimal (r, q) policy in a perfect manner (Browne and Zipkin [20]). As a result, several algorithms are developed. Federgruen and Zheng [39] developed an algorithm for solving a model similar to that of Hadley and Whitin’s [51]. Matheus and Gelders [77] considered an (r,q) inventory system that supports a number of customers whose demand pattern is compound Poisson. By varying the reorder point while keeping a constant order quantity, they could use the model to cope with different desired service levels . Moinzadeh and Nahmias [81] employed two (r,q) policies to handle a single product, from which one ordering strategy is for emergency purpose. On the other hand, Moinzadeh and Lee [80] considered a two shipments policy in which an order q is partially shipped in two different time units. Again, they used EOQ as a lower bound of the order quantity and they formulated a searching algorithm for the reorder point r, given the value q. They considered both Poisson and normal demands. For more dynamic systems, Badinelli [8], Axsäter [7] and Ng et al. [84] proposed some (r,q) policies that could be applied to an inventory system with more than one levels or facilities. Axsäter’s model [7] assumed that the inventory positions at all retailers are uniformly distributed, while Ng et al.’s model [84] assumed that they followed a Poisson distribution .

1.4.3 Markov Chain

Browne and Zipkin [20] developed an (r, q) policy in which the demand is a time-homogeneous Markov process. They designed an algorithm to evaluate the policy. In their model, they assumed the inventory position is uniformly distributed in the interval \((r,r+q)\). Parlar and Perry’s Markovian model is used to tackle supply uncertainty [92]. They checked the availability of the supply before placing an order of quantity q when inventory level drops to a reorder point r. If there is enough stock from the supplier, replenishment is made in zero lead-time . This state is called an “ON” state. If the stock is not available, then the order will arrive after T periods. It is called an “OFF” state. They tried to find an optimal (r,q,T) policy for minimizing a long-run average cost function based on the renewal reward theorem. Melchiors [79] applied Markov chain in comparing two can-order policies suggested by two other authors. He considered twelve products whose demands are Poisson and used simulation in comparison. Despite of coping with inventory management problems, Markov chain analysis is widely used in modelling machine breakdowns (Abboud [1]), soil conditions testing (Taha [111]), accounts receivable system (Render et al. [98]) and so on.

1.5 Concluding Remarks

Undoubtedly, manufacturing global products with customization for local markets plays an important role in striving for a distinctive competitive advantage for players in a global supply chain. Postponement enables companies to achieve higher manufacturing flexibility and product quality at lower costs. Pull postponement makes use of customer demand information to determine the push–pull boundary in a supply chain that allows common processes to be completed before those that differentiate product characteristics. Logistics postponement aims at deciding whether pull postponement should be carried out at local facilities instead of in the central production line. Form postponement is an enabling strategy for pull postponement as it opts for the use of standardized components and processes to achieve customization. Price postponement is an economic strategy that resorts to postponing the setting of product price. Practically, they can be combined and applied simultaneously to achieve optimization in a supply chain.

As mentioned, postponement is not a panacea to all situations. Examples cited in this chapter are on a situational basis. As a matter of fact, more generalized models and frameworks need to be developed. They can offer better insights and supportive evidence for postponement implementation in different areas. In this book, four types of model are presented to evaluate the impacts of pull and form postponement strategies under various supply chain structures. First, we develop two EOQ-based models to examine the impact of pull postponement in  Chapter 2. Then we develop some EPQ-based models to examine the impact of postponement in  Chapter 3. In  Chapter 4 we propose a stochastic model of a single end-product supply chain that consists of a supplier, a manufacturer and a number of customers. In  Chapter 5 we aim at conducting a simulation experiment of a two-end-product supply chain, for which customer demands are discrete and independent. Besides mathematical models, two case studies from industry are presented to support our theoretical results in  Chapter 6. In  Chapter 7 we conclude the book and suggest some worthy topics for future research.


  1. 1.

    The case study is presented in  Chapter 6.


  1. 3.
    Alderson, W. 1957, Marketing efficiency and the principle of postponement, Marketing Behavior and Executive Action, Richard D. Irwin, Homewood, IL pp. 423–427.Google Scholar
  2. 43.
    Garg, A. and Lee, H. L. 1998, Managing product variety: an operations perspective, in Tayur, S., Ganesham, R. and Magazine, M. (editors), Quantitative Models for Supply Chain Management, Kluwer Academic Publishers, Boston/Dordrecht/London.Google Scholar
  3. 41.
    Feitzinger, E. and Lee, H. L. 1997, Mass customization at Hewlett-Packard: the power of postponement, Harvard Business Review, January–February, 116–121.Google Scholar
  4. 73.
    Li, S., Ragu-Nathan, B., Ragu-Nathan, T. S. and Rao, S. S. 2006, The impact of supply chain management practices on competitive advantage and organizational performance, OMEGA, the International Journal of Management Science, 34, 2, 107–124.CrossRefGoogle Scholar
  5. 75.
    MacDuffie, J. P., Sethuraman, K. and Fisher, M. L. 1996, Product variety and manufacturing performance: evidence from the international automotive assembly plant study, Management Science, 42, 3, 350–369.CrossRefGoogle Scholar
  6. 128.
    Yeh, K. H. and Chu, C. H. 1991, Adaptive strategies for coping with product variety decisions, International Journal of Operations and Production Management, 11, 8, 35–47.CrossRefGoogle Scholar
  7. 94.
    Porter, M. 1980, Competitive Strategy: Techniques for Analyzing Industries and Competitors, The Free Press, New York.Google Scholar
  8. 60.
    Kahn, B. E. 1998, Dynamic relationship with customers: high-variety strategies, Academy of Marketing Science Journal, 26, 1, 45–53.CrossRefGoogle Scholar
  9. 109.
    Stalk, G. Jr. 1988, Time-the next source of competitive advantage, Harvard Business Review, 66, 4, 45–51.Google Scholar
  10. 115.
    Ulrich, K., Randall, T., Fisher, M. and Reibstein, D. 1999, Managing product variety, in Ho, T. H. and Tang, C. S. (editors), Product Variety Management: Research Advances, Kluwer Academic Publishers, Boston/Dordrecht/London.Google Scholar
  11. 76.
    Martin, M., Hausman, W. and Ishii, K. 1999, Design for variety, in Ho, T. H. and Tang, C. S. (editors), Product Variety Management: Research Advances, Kluwer Academic Publishers, Boston/Dordrecht/London.Google Scholar
  12. 63.
    Lampel, L. and Mintzberg, H. 1996, Customizing customization, Sloan Management Review, 38, 21–30.Google Scholar
  13. 65.
    Lee, H. L. 1998, Postponement for mass customization: satisfying customer demands for tailor-made products, in Gattorna, J. (editor), Strategic Supply Chain Alignment, Gower, England.Google Scholar
  14. 70.
    Lee, H. L. and Tang, C. S. 1998, Variability reduction through operations reversal, Management Science, 44, 2, 162–172.CrossRefGoogle Scholar
  15. 64.
    Lee, H. L. 1996, Effective inventory and service management through product and process redesign, Operations Research, 44, 1, 151–159.CrossRefGoogle Scholar
  16. 29.
    Christopher, M. 2000, The agile supply chain: competing in volatile markets, Industrial Marketing Management, 29, 1, 37–44.CrossRefGoogle Scholar
  17. 93.
    Pine II, B. J. 1993, Mass Customization: The New Frontier in Business Competition, Harvard Business School Press, Boston.Google Scholar
  18. 40.
    Federgruen, A. and Zipkin, P. 1984, Approximations of dynamic multi-location production and inventory problems, Management Science, 30, 69–84.CrossRefGoogle Scholar
  19. 105.
    Simchi-Levi, D., Kaminsky, P. and Simchi-Levi, E. 2000, Designing and Managing the Supply Chain, Concepts, Strategies, and Case Studies, McGraw-Hill, New York.Google Scholar
  20. 124.
    van Mieghem, J. A. and Dada, M. 1999, Price versus production postponement: capacity and competition, Management Science, 45, 12, 1631–1649.Google Scholar
  21. 101.
    Schroeder, R. 1993, Operations Management: Decision Making in the Operations Function, 4th Edition, McGraw-Hill, New York.Google Scholar
  22. 19.
    Browne, J., Harhen, J. and Shivnan, J. 1996, Production Management Systems: An Integrated Perspective, 2nd Edition, Addision-Wesley, Harlow, England.Google Scholar
  23. 5.
    Arnold, J. 1998, Introduction to Materials Management, 3rd Edition, Prentice-Hall, Upper Saddle River, New Jersey.Google Scholar
  24. 68.
    Lee, H. L., Padmanabhan, V. and Whang, S. 1997, The bullwhip effect in supply chain, Sloan Management Review, 38, 3, 93–102.Google Scholar
  25. 35.
    De Hann, J., Yamamoto, M. and Lovink, G. 2001, Production planning in Japan: rediscovering lost experiences or new insight? International Journal of Production Economics, 71, 1–3, 101–109.CrossRefGoogle Scholar
  26. 55.
    Hoekstra, S. J. and Romme, J. H. J. M. 1992, Integral Logistics Structures: developing Customer-Oriented Goods Flow, McGraw-Hill, London.Google Scholar
  27. 17.
    Brown, A. O., Lee, H. L. and Petrakian, R. 2000, Xilinx improves its semiconductor supply chain using product and process postponement, Interfaces, 30, 4, 65–80.CrossRefGoogle Scholar
  28. 117.
    van Donk, D. P. 2001, Make to stock or make to order: the decoupling point in the food processing industries, International Journal of Production Economics, 69, 297–306.CrossRefGoogle Scholar
  29. 118.
    van Hoek, R. 1997, Postponed manufacturing: a case study in the food supply chain, Supply Chain Management, 2, 2, 63–75.CrossRefGoogle Scholar
  30. 21.
    Bucklin, L. P. 1965, Postponement, speculation and the structure of distribution channels, Journal of Marketing Research, February, 26–31.Google Scholar
  31. 116.
    van Dijk, S., van der Vorst, J. and Beulens, A. 2003, Chicken run-the poultry supply chain, in Johnston, R., Chambers, S., Harland, C., Harrison, A. and Slack, N. (editors), Cases in Operations Management, 3rd Edition, Prentice Hall, UK.Google Scholar
  32. 14.
    Bowersox, D. J. and Closs, D. J. 1996, Logistical Management, the Integrated Supply Chain Process, Macmillan Publishing Company, New York.Google Scholar
  33. 67.
    Lee, H. L., Billington, C. and Carter, B. 1993, Hewlett-Packard gains control of inventory and service through design for localization, Interfaces, 23, 4, 1–11.CrossRefGoogle Scholar
  34. 114.
    Twede, D., Clarke, R. H. and Tait, J. A. 2000, Packaging postponement: a global packaging strategy, Packaging Technology and Science, 13, 3, 105–115.CrossRefGoogle Scholar
  35. 2.
    Ackerman, K. B. 1997, Postponement, Practical Handbook of Warehousing, 4th Edition, Chapman and Hall, New York.Google Scholar
  36. 18.
    Brown, G., Keegan, J., Vigus, B. and Wood, K. 2001, The Kellogg company optimizes production, inventory, and distribution, Interfaces, 31, 6, 1–15.Google Scholar
  37. 108.
    South China Morning Post (SCMP), 2002, Institutions Happy with Price of BoC, 23 July 2002.Google Scholar
  38. 107.
    South China Morning Post (SCMP), 2002, BoC Pledges 500 Shares to Every Applicant, 22 July 2002.Google Scholar
  39. 50.
    Graman, G. A. and Magazine, M. J. 2002, A numerical analysis of capacitated postponement, Production and Operations Management, 11, 3, 340–357.CrossRefGoogle Scholar
  40. 27.
    Cheng, T. C. E. and Podolsky, S. 1996, Just-in-time Manufacturing: An Introduction, 2nd Edition, Chapman and Hall, London.Google Scholar
  41. 52.
    Heskett, J. L. 1976, Marketing, Macmillan Publishing Co., Inc., New York.Google Scholar
  42. 38.
    Ernst, R. and Kamrad, B. 2000, Evaluation of supply chain structures through modularization and postponement, European Journal of Operational Research, 124, 495–510.CrossRefGoogle Scholar
  43. 89.
    Olhager, J. 1994, On the positioning of the customer order decoupling point, Proceedings from the 1994 Pacific Conference on Manufacturing, Jakarta, pp. 1093–1100.Google Scholar
  44. 54.
    Hoekstra, S. J. and Romme, J. H. J. M. 1985, Direction Towards Integral Logistics, Kluwer, Deventer, The Netherlands.Google Scholar
  45. 28.
    Chiou, J. S., Wu, L. Y. and Hsu, J. C. 2002, The adoption of form postponement strategy in a global logistics system: the case of Taiwanese information technology industry, Journal of Business Logistics, 23, 107–124.CrossRefGoogle Scholar
  46. 120.
    van Hoek, R. I. 2000, Role of third party logistics services in customization through postponement, International Journal of Service Industry Management, 11, 4, 374–387.CrossRefGoogle Scholar
  47. 113.
    Tully, S. 1993, The module corporation, Fortune, 8 February, 52–56.Google Scholar
  48. 74.
    Ma, S., Wang, W. and Liu, L. 2002, Commonality and postponement in multistage assembly systems, European Journal of Operational Research, 142, 3, 523–538.CrossRefGoogle Scholar
  49. 130.
    Zinn, W. and Bowersox, D. J. 1988, Planning physical distribution with the principle of postponement, Journal of Business Logistics, 9, 2, 117–136.Google Scholar
  50. 62.
    Kennedy, D., Elliott, B. and Carbone, F. 1998, Lean synchronous manufacturing: “hard-wiring” production to demand, in Gattorna, J. (editor), Strategic Supply Chain Alignment, Gower, England.Google Scholar
  51. 11.
    Barkan, P. 1991, Strategic and tactical benefits of simultaneous engineering, Design Management Journal, spring, 39–42.Google Scholar
  52. 23.
    Calvin III, R. K. and Miller, W. H. 1989, Design for manufacturing, SRC Newsletter, 7, 1–3.Google Scholar
  53. 123.
    van Hoek, R. I., Vos, B. and Commandeur, H. R. 1999, Restructuring European supply chains by implementing postponement strategies, Long Range Planning, 32, 5, 505–518.CrossRefGoogle Scholar
  54. 12.
    Beamon, B. M. 1998, Supply chain design and analysis: models and methods, International Journal of Production Economics, 55, 3, 281–294.CrossRefGoogle Scholar
  55. 53.
    Hillier, F. S. and Lieberman, G. J. 2001, Introduction to Operations Research, 7th Edition, McGraw-Hill, Singapore.Google Scholar
  56. 69.
    Lee, H. L. and Tang, C. S. 1997, Modeling the costs and benefits of delayed product differentiation, Management Science, 43, 1, 40–53.CrossRefGoogle Scholar
  57. 44.
    Garg, A. and Tang, C. S. 1997, On postponement strategies for product families with multiple points of differentiation, IIE Transactions, 29, 8, 641–650.Google Scholar
  58. 6.
    Aviv, Y. and Federgruen, A. 2001, Capacitated multi-item inventory systems with random and seasonally fluctuating demands: implications for postponement strategies, Management Science, 47, 4, 512–531.CrossRefGoogle Scholar
  59. 32.
    Collier, D. A. 1982, Aggregate safety stock levels and component part commonality, Management Science, 28, 11, 1296–1303.CrossRefGoogle Scholar
  60. 78.
    McClain, J. O., Maxwell, W. L., Muckstadt, J. A., Thomas, L. J. and Weiss, E. N. 1984, Comment on “Aggregate safety stock levels and component part commonality”, Management Science, 30, 6, 772–774.CrossRefGoogle Scholar
  61. 9.
    Baker, K. R., Magazine, J. and Nuttle, H. L. W. 1986, The effect of commonality on safety stock in a simple inventory model, Management Science, 32, 8, 982–988.CrossRefGoogle Scholar
  62. 10.
    Ballou, R. H. 1989, Heuristics: rule of thumb for logistics decision making, Journal of Business Logistics, 10, 1, 122–132.Google Scholar
  63. 33.
    Corry, P. and Kozan, E. 2002, Optimizing a horizontally integrated push/pull hybrid production system in a foundry, in Kozan, E. and Ohuchi, A. (editors), Operations Search/Management Science at Work, Kluwer Academic Publishers, Boston/Dordrecht/London.Google Scholar
  64. 31.
    Cochran, J. K. and Kim, S. S. Optimal junction point location and inventory levels in serial hybrid push/pull production system, International Journal of Production Research, 36, 4, 1141–1155.Google Scholar
  65. 129.
    Zinn, W. 1990, Developing heuristics to estimate the impact of postponement on safety stock, The International Journal of Logistics Management, 1, 2, 11–16.CrossRefGoogle Scholar
  66. 59.
    Johnson, M. E. and Anderson, E. 2000, Postponement strategies for channel derivatives, The International Journal of Logistics Management, 11, 1, 19–35.CrossRefGoogle Scholar
  67. 121.
    van Hoek, R. I. 2000, The role of third-party logistics providers in mass customization, International Journal of Logistics Management, 11, 37–46.CrossRefGoogle Scholar
  68. 56.
    Huang, C. W. and Lo, C. P. 2003, Using postponed manufacturing to reconfigure the supply chain in the desktop personal company industry: the case of Taiwan, International Journal of Management, 20, 241–256.Google Scholar
  69. 57.
    Jahre, M. 1995, Household waste collection as a reverse channel: a theoretical perspective, International Journal of Physical Distribution and Logistics Management, 25, 2, 39–55.CrossRefGoogle Scholar
  70. 85.
    Nicoll, A. D. 1994, Integrating logistics strategies, Annual International Conference Proceedings-American Production and Inventory Control Society, pp. 590–594.Google Scholar
  71. 125.
    Wan, J. C. L. 2004, Postponement Strategy in Supply Chain Management (M.Phil. Thesis), Department of Logistics, The Hong Kong Polytechnic University, Hong Kong.Google Scholar
  72. 71.
    Li, J., Cheng, T. C. E. and Wang, S. Y. 2007, Analysis of postponement strategy for perishable items by EOQ-based models, International Journal of Production Economics, 107, 1, 31–38.CrossRefGoogle Scholar
  73. 72.
    Li, J., Wang, S. Y. and Cheng, T. C. E. 2008, Analysis of postponement strategy by EPQ-based models with planned backorder, OMEGA, the International Journal of Management Science, 36, 5, 777–788.CrossRefGoogle Scholar
  74. 119.
    van Hoek, R. I. 1998, Logistics and virtual integration: postponement, outsourcing and the flow of information, International Journal of Physical Distribution and Logistics Management, 28, 7, 508–523.CrossRefGoogle Scholar
  75. 131.
    Zipkin, P. H. 2000, Foundations of Inventory Management, International Edition, McGraw-Hill, Boston.Google Scholar
  76. 51.
    Hadley, G. and Whitin, T. M. 1963, Analysis of Inventory Systems, Prentice-Hall, Englewood Cliffs, New Jersey.Google Scholar
  77. 104.
    Silver, E. A. 1976, A simple method of determining order quantities in joint replenishments under deterministic demand, Management Science, 22, 12, 1351–1361.CrossRefGoogle Scholar
  78. 103.
    Shu, F. T. 1971, Economic ordering frequency for two items jointly replenished, Management Science, 17, 6, 406–410.CrossRefGoogle Scholar
  79. 86.
    Nocturne, D. J. 1973, Economic ordering frequency for several items jointly replenished, Management Science, 19, 9, 1093–1096.CrossRefGoogle Scholar
  80. 25.
    Chen, C. K. and Min, K. J. 1994, A multi-product EOQ model with pricing consideration-T. C. E. Cheng’s model revisited, Computers and Industrial Engineering, 26, 4, 787–794.CrossRefGoogle Scholar
  81. 26.
    Cheng, T. C. E. 1990, An EOQ model with pricing consideration. Computers and Industrial Engineering, 18, 4, 529–534.CrossRefGoogle Scholar
  82. 83.
    Nahmias, S. 2001, Production and Operations Analysis, 4th Edition, McGraw-Hill International Edition, New York.Google Scholar
  83. 37.
    Eilon, S. 1985, Multi-product batch production on a single machine-a problem revisited, OMEGA, the International Journal of Management Science, 13, 5, 453–468.CrossRefGoogle Scholar
  84. 46.
    Goyal, S. K. 1984, Determination of economic production quantities for a two-product single machine system, International Journal of Production Research, 22, 1, 121–126.CrossRefGoogle Scholar
  85. 112.
    Tersine, R. J. 1994, Principles of Inventory and Material Management, 4th Edition, Prentice-Hall, New Jersey.Google Scholar
  86. 22.
    Byrne, M. D. 1990, Multi-item production lot sizing using a search simulation approach, Engineering Costs and Production Economics, 19, 1–3, 307–311.CrossRefGoogle Scholar
  87. 87.
    O’Grady, P. J. and Byrne, M. D. 1987, Lot sizing in a practical manufacturing system, Engineering Costs and Production Economics, 13, 1, 13–26.CrossRefGoogle Scholar
  88. 88.
    O’Grady, P. and Byrne, M. 1989, Adaptive lot sizing (with case study), Engineering Costs and Production Economics, 15, 397–404.CrossRefGoogle Scholar
  89. 13.
    Bertrand, J. W. M. 1985, Multiproduct optimal batch sizes with in-process inventories and multi work centers, IIE Transactions, 17, 2, 157–163.CrossRefGoogle Scholar
  90. 49.
    Goyal, S. K., Gunasekaran, A., Martikainen, T. and Yli-Olli, P. 1993, Determining economic batch sizes for multiple items in a multilevel manufacturing system, International Journal of Operations and Production Management, 13, 12, 79–93.CrossRefGoogle Scholar
  91. 39.
    Federgruen, A. and Zheng, Y. S. 1992, An efficient algorithm for computing an optimal (r,q) policy in continuous review stochastic inventory systems, Operations Research, 40, 808–813.CrossRefGoogle Scholar
  92. 20.
    Browne, S. and Zipkin, P. 1991, Inventory models with continuous stochastic demands, The Annals of Applied Probability, 1, 419–435.CrossRefGoogle Scholar
  93. 77.
    Matheus, P. and Gelders, L. 2000, The \((R,Q)\) inventory policy subject to a compound Poisson demand pattern, International Journal of Production Economics, 68, 3, 307–317.CrossRefGoogle Scholar
  94. 81.
    Moinzadeh, K. and Nahmias, S. 1988, A continuous review model for an inventory system with two supply modes, Management Science, 34, 761–773.CrossRefGoogle Scholar
  95. 80.
    Moinzadeh, K. and Lee, H. L. 1989, Approximate order quantities and reorder points for inventory systems where orders arrive in two shipments, Operations Research, 37, 277–287.CrossRefGoogle Scholar
  96. 7.
    Axsäter, S. 1997, Simple evaluation of echelon stock \((R,Q)\) policies for two-level inventory systems, IIE Transactions, 29, 661–669.Google Scholar
  97. 8.
    Badinelli, R. D. 1992, A model for continuous-review pull policies in serial inventory systems, Operations Research, 40, 142–156.CrossRefGoogle Scholar
  98. 84.
    Ng, C. T., Li, L. Y. O. and Chakhlevitch, K. 2001, Coordinated replenishments with alternative supply sources in two-level supply chains, International Journal of Production Economics, 73, 227–240.CrossRefGoogle Scholar
  99. 92.
    Parlar, M. and Perry, D. 1995, Analysis of a \((Q,r,T)\) inventory policy with deterministic and random yields when future supply is uncertain, European Journal of Operational Research, 84, 431–443.CrossRefGoogle Scholar
  100. 79.
    Melchiors, P. 2002, Calculating can-order policies for the joint replenishment problem by the compensation approach, European Journal of Operational Research, 141, 587–595.CrossRefGoogle Scholar
  101. 1.
    Abboud, N. E. 2001, A discrete-time Markov production-inventory model with machine breakdowns, Computers and Industrial Engineering, 39, 95–107.CrossRefGoogle Scholar
  102. 111.
    Taha, H. A. 2003, Operations Research: An introduction, 7th Edition, Prentice-Hall, New Jersey.Google Scholar
  103. 98.
    Render, B., Stair, R. M. Jr. and Hanna, M. 2003, Quantitative Analysis for Management, Prentice-Hall, New Jersey.Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • T.C. Edwin Cheng
    • 1
    Email author
  • Jian Li
    • 2
  • C.L. Johnny Wan
    • 1
  • Shouyang Wang
    • 3
  1. 1.Department of Logistics & Maritime StudiesThe Hong Kong Polytechnic UniversityKowloonHong Kong SAR
  2. 2.School of Economics & Management Beijing University of Chemical Technology (BUCT)BeijingChina, People’s Republic
  3. 3.Chinese Academy of Sciences Academy of Mathematics & Systems ScienceBeijingChina, People’s Republic

Personalised recommendations