ظرفیت مناسب: ساخت سوداگرانه و واکنش پذیر محصولات مد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|10750||2008||15 صفحه PDF||سفارش دهید||9467 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Production Economics, Volume 114, Issue 2, August 2008, Pages 416–430
Consider a fashion goods retailer choosing a strategy for contracting production of its products. It can (1) speculate by contracting for a certain quantity to be produced well ahead of uncertain demand at relatively low unit cost, (2) react by waiting until demand is known, and only then contracting for just the right quantity at a higher unit cost, or (3) hedge its bets by speculating on a portion of the total quantity, and reacting to demand for the rest. Using a two-product two-stage model, we identify the conditions under which each strategy is preferred, and determine capacity requirements. We find that fashion retailing often benefits from the dual strategy due to relatively higher obsolescence costs. But the use of the dual strategy is sensitive to the cost premium for reactive capacity and to the makeup of reactive production costs as either largely variable or fixed.
Consider SportTee, a hypothetical fashion goods retailer that sells three different lines of college and professional team logo shirts: (1) TeamTees: official team jerseys with soccer, football, baseball, or basketball insignias; (2) ChampionTees: “event” shirts sold immediately after team victories in the World Cup, NCAA championship, SuperBowl, or World Series; and (3) PlayerTees: featuring specific player names and their numbers, such as soccer jerseys for Los Angeles Galaxy David Beckham and for the Salt Lake Real's Freddie Adu. How should SportTee tailor production contract strategies to best fit the unique characteristics of each business? Fisher and Raman (1996) proposed the strategy of accurate response as a means of better matching supply with demand in the fashion goods industry. Early in the production season, prior to realization of demand, capacity is dedicated to products based on a forecast, and production is referred to as speculative. Later in the season, demand is essentially known and production is reactive, in that it responds to demand volume that is known with more certainty. Fisher and Raman determine the production sequence for a set of known products given a fixed amount of speculative and reactive capacity (i.e., they establish which products should be manufactured at each of the two stages of the season given that capacities cannot be changed). We focus on the capacity decision in a setting similar to that of Fisher and Raman. In particular, we seek insight into how much speculative volume to produce prior to realizing demand and how much reactive capacity to have available for production in response to realized demand. Speculative production typically is more efficient because, for example, the firm may have the opportunity to mass-produce the items in a low-cost environment. A mix of speculative and reactive capacity (as in Fisher and Raman) is one strategy, but what is the right mix of speculative and reactive capacity under this strategy, and are there times when a firm might use a pure speculative strategy, or a pure reactive strategy? We develop a simple two-stage, single-period model that determines the optimal speculative and reactive capacity levels, first for the case of one product, and then for the case of two products with independent and identically distributed demands. In the first stage, the firm commits to its speculative and reactive capacity levels, incurring fixed costs for each, and produces to the speculative capacity level. In the second stage, we assume that uncertainty is fully resolved—a slightly stronger assumption than Fisher and Raman make—and the firm uses its reactive capacity to fill this demand realization. If the sum of speculative and reactive capacity is insufficient, the firm experiences lost sales, representing an underage cost. It experiences an obsolescence (overage) cost if demand falls short of speculative production. If demand falls between the speculative capacity and total capacity levels, then some of the (relatively costly) reactive capacity goes unused. In the two-product case, reactive capacity can be used to fill demand for either product. Returning to the SportTee example, a season's demand for a TeamTee (e.g., a Dallas Cowboys shirt), might be largely predictable based on historic patterns over prior seasons, and thus it might be optimal to make all units of this product using capacity that is speculative. In contrast, demand for ChampionTees with insignias such as “Colts: 2007 AFC Champions” and “Bears: 2007 NFC Champions” can only be determined after the outcome of the championship match is observed, such that it might be optimal to make all units using reactive capacity. And since the 2007 seasonal demand for the Beckham #23 Los Angeles Galaxy shirt or the Freddie Adu shirt may be associated with a high variance, it may be optimal for this product to use a mix of speculative and reactive capacities. Our framework applies to a scenario where the retailer has the opportunity to contract for the production of some units of a specific product prior to its release into the market and then has a chance to supplement this initial run (of speculative production) with one additional run (that uses reactive capacity) after getting the actual orders for the product. While the primary focus of this paper is on the fashion clothing industry (see our examples from Sport Obermeyer later in the paper), other retailers and manufacturers face similar types of capacity decisions (that is, products other than clothing can also be considered to be “fashion goods”). For example, product life cycles in the personal computer (PC) industry have become sufficiently short that PC manufacturers may initiate a very limited number of production runs of any specific model before moving on to the next model. Initial runs might use speculative capacity followed by later runs using reactive capacity. Consider also a semiconductor manufacturer who would like to take advantage of low-cost, long lead-time, offshore production. To the extent that long lead-times force the firm to set the offshore production quantity before observing any demand, that capacity is speculative. The manufacturer might benefit by reserving additional domestic reactive capacity (presumably at a higher cost than the speculative capacity) so that it could respond to the additional demand information after getting an early demand signal. We model the case where the product life cycle permits only one chance for recourse after demand becomes known. Finally, consider the automobile industry, which is also facing shorter model lifetimes. Prior to the introduction of a new model, the firm might have to commit to a production schedule that is difficult to change significantly over the product's lifetime. Physical constraints may limit increases in capacity while labor agreements and supplier contracts may restrict decreases. As a result, the auto manufacturer may need to pre-commit to a production rate that effectively represents speculative capacity. Simultaneously, the firm could develop a contingency plan for producing additional units in case strong demand materializes, using some form of reactive capacity. For example, the firm might reserve some production slots in a second plant that is more flexible. Our model lends insight into the tradeoff between these speculative and reactive capacity commitments. These examples illustrate settings where the firm may want to adopt a dual strategy that utilizes both speculative and reactive capacities. The firm's initial production comes from efficient speculative capacity (e.g. lower-cost, mass-produced T-shirts made offshore with longer lead times). But since speculative output is unlikely to match demand exactly, the firm limits its speculative output, while holding some reactive capacity (e.g. higher cost, locally manufactured T-shirts produced quickly after demand is observed) that it uses only if demand exceeds speculative capacity. With our single-period two-stage model, we first assume there is a single-product. We identify the cases where the firm employs (1) only speculative capacity, (2) only reactive capacity, or (3) both speculative and reactive capacity. We find newsvendor-type results describing the optimal levels of speculative and reactive capacities the firm should hold and offer an interpretation of underage and overage costs. We then consider the two-product situation where the flexible reactive capacity can be used to fill demand for either product. Again, we identify the boundaries between the three cases and find newsvendor-type results describing capacity levels. In both the single-product and the two-product cases, we find that the dual approach is most attractive when obsolescence costs are high (i.e., when salvage value is low), and when the premium for reactive production is relatively low. With the two-product case, reactive capacity offers the opportunity to pool demands and thus increases the attractiveness of reactive capacity relative to the one-product case. Generally, pooling results in lower resource requirements (such as holding less inventory). Surprisingly, we find the firm may increase its overall capacity resource level when the flexible reactive capacity can fill demands for two products, as compared to a situation where reactive capacity applies only to one product. While others have noted that pooling can lead to increased inventory for better demand response (for example, see Gerchak and Mossman, 1992), our framework results in increased capacity for improved reaction to demand. The intuition is that the reactive capacity becomes more valuable in the two-product case, so the firm may choose to hold more. The paper is organized as follows. In Section 2, we review related research. In Section 3, we introduce the model and develop the results for the case of a single product. In Section 4, we extend the model to two products and illustrate model results given three contrasting firms. We offer a discussion of results and conclusion in Section 5.
نتیجه گیری انگلیسی
We have examined how a firm might approach its capacity acquisition decisions in a setting similar to the ‘‘accurate response’’ situation analyzed by Fisher and Raman. While they assume the firm has an infinite level of speculative capacity and a pre-determined (or set) capacity level for reactive production, we analyze the case where the two capacity levels are endogenous with some simplifying assumptions. Our results suggest the firm may want to acquire capacity of only one type, or of both types, depending on the cost premium for reactive capacity. If the reactive capacity costs no more than the speculative type, then the firm uses only reactive capacity; it uses some of each type unless the cost premium for the reactive type is ‘‘too high,’’ in which case only the speculative type is used. While we make no formal assumption with regard to the relative costs of speculative and reactive capacities, it seems likely that reactive production capacity will be more expensive. In Fisher and Raman, the only explicit difference between the two types of capacity is in the timing of production—the firm operates a plant of a given fixed size nearly year-round, using that capacity in speculative fashion until a strong demand signal is received at its spring trade show, after which time the capacity represents the reactive type in that it can be used to respond to the better demand information. Thus, the cost difference between the two capacity types might appear to be zero, even though the value of reactive capacity to the firm clearly exceeds that of speculative capacity. When capacity is considered endogenously, it is likely that production capacity purchased in an open market would be priced differently depending on its ability to be flexible and reactive and depending on the time of the year. One can imagine other settings where speculative and reactive capacities are physically different, instead of differing in timing of production. For example, by pre-building items in speculative fashion, a firm can benefit from steady material flows that facilitate fewer setups and less complexity. On the other hand, reactive production possibly requires more sophisticated resources and plant flexibility, particularly as related to our two-product case. Given that the cost difference between the two capacity types determines the optimal level of each,initiatives such as flexible manufacturing, more efficient shop floor control resulting from more intense information processing, information sharing up and down the supply chain, and other supply chain improvements that reduce the cost of reactive capacity will give firms currently using a strict speculative approach an incentive to switch to a dual speculative/reactive strategy or even a strict reactive strategy.In addition to using the case of Sport Obermeyer to illustrate application to the fashion clothing retail industry, we discussed in the Introduction a hypothetical company, SportTee. In real life, Reebok faces related issues, as discussed in a case study by Parsons and Graves (2005). They use a combination of speculative and reactive capacity to produce team and player jerseys for the National Football League. The firm uses purely speculative capacity to make nameless team jerseys. For players of lesser popularity, the firm uses purely reactive capacity to sew the names on the jerseys in the USA. For the most popular players, speculative capacity is used to sew the player’s name on a fraction of jerseys in the Far East, followed by the use of reactive capacity to fill any remaining demand.Future work could address implications of a dual strategy approach not considered in our simplified setting. For example, we do not consider the marketing implications of reactive capacity. Customers may be willing to pay more for products customized to their specifications, made possible with the use of reactive capacity, as opposed to buying an off—theshelf product that may contain unwanted features or lack other desired features. The National Bicycle Industrial Company, as described by Fisher (1994), effectively uses reactive production for lower-volume, up-scale bikes, while sticking with speculative production for the more basic models. Thus the dual strategy expands its product line toward the upper end, and has significant implications on both overall product variety and product demand. At the same time there are potential negative market aspects of reactive production, such as customer impatience with possible delays in receiving goods produced reactively.The idea of postponement, as discussed by Lee and Tang (1997), is related to our notion of dual speculative/reactive capacities: postponement effectively involves initiating production of a given item using speculative production, and then completing it using reactive capacity after demand is realized. An extension of our model might examine whether postponement is preferred to the dual strategy that we analyze, or whether some combination of postponement and our dual strategy performs better than either methodology used individually.Extensions of our model to the case of correlated demand, and to more complicated cost structures,such as economies or diseconomies of scale, are other possible future research directions.We suspect that the results in Bengtsson and Olhager (2002),developed from option-pricing theory, would hold directionally in our setting, in that the value of flexibility increases with the number of products and decreases with more positively correlated demand. In summary, our model lends some understanding of the factors pushing the firm toward strictly speculative capacity, strictly reactive capacity, or a dual strategy approach that uses both. We have provided insights into the impact of product obsolescence costs, capacity costs, and the cost premium of flexible reactive production on the capacity choice. Our results suggest that by considering these factors in setting speculative capacity and reactive capacity levels, the firm can cope more profitably with uncertain demand.