Knowledge that Transforms

wepublished the extended abstracts of the 1999 winners in the hopes that this could becomean annual event. Our hopes have become a reality.Sridhar Seshadri and Ravi Anupindi, New York University, cochaired the 2000 competi-tion. The judges for the final round were Garrett van Ryzin, Columbia University; YehudaBassok, University of Southern California; and Michael Pinedo, New York University.The first-prize winner received $400, while second prize received $200. Prizes were award-ed at the INFORMS meeting in San Antonio, Texas in November 2000. All finalists receiveda $50.00 discount coupon redeemable at an ‘‘e-tailer.’’The winners and their faculty mentors were:

We characterize optimal policies for the problem of allocating a single server to a set of jobs from N families. Each job is an instance of demand for an item and is associated with a family, a holding cost rate, and a mean processing time. Set-up times are required to switch from one family to another, but are not required to switch within a family. We consider the case in which the order of jobs within the family is unconstrained, and a variation in which the order is fixed. The optimization is with respect to the weighted flowtime, and we treat problems both with and without a makespan-constraint. Practical examples based on this model are described. We partially characterize an optimal policy by means of a Gittins rewardrate index and a similar switching index derived from multi-armed bandit theory. For deterministic problems with a makespan constraint, we present an optimization algorithm for the special case of two families and at most three set-ups . Without a makespan constraint and without preemption, we prove that our analysis of a deterministic model extends to stochastic set-up and processing times without loss of optimality. Managerial insights based on our technical results are provided.

Demand forecasts do not become consistently more accurate as they are updated. We present examples demonstrating this counterintuitive phenomenon and some theoretical results to explain its occurrence. Specifically, we analyze the effect of demand randomness on forecast-update performance. A surprising result is that under various theoretical models involving demand randomness alone, updated forecasts will be less accurate between 30% and 50% of the time.

The third-party logistics industry has grown rapidly in recent years, accounting for $46 billion of the total $921 billion in logistics spending in the United States during 1999. This figure is expected to grow by 15 to 20% annually as manufacturing firms increasingly partner with third-party logistics providers to cost effectively distribute their products, while meeting increasingly stringent service expectations of customers. These logistics partnerships have introduced a new set of decision requirements to negotiate compensation for distribution services. Based on a collaborative project with a leading building products manufacturer, this paper describes the development and implementation of a novel linear programming model to decide the delivery fees paid to distributors. The model applies to manufacturer-distributor partnerships where distributors are compensated using fee values that depend on delivery weights and distances. It ensures that the expected compensation, considering stochastic demands, is adequate to cover the aggregate distribution costs for each distributor, and permits imposing various consistency conditions to ensure that fee values are credible. The model proved effective in helping the manufacturer develop a new fee table that generated considerable economic savings and provided more equitable compensation to distributors.

Some firms, particularly in high-tech, appear to view technology leadership and cost leadership as separate and distinct ways of achieving high profits within a given product market. In contrast, we develop a model of technology competition suggesting that a firm's success in sustaining its technology leadership may hinge on its ability to produce the new product (resulting from the new technology) at lower cost, an ability we call the firm's “cost competence.” In our model, cost competence, above a critical hurdle level, is essential to the incumbent firm in its bid to retain technology leadership. The model also clarifies the role of “innovative competence,” which characterizes a firm's ability to turn an investment in new technology into a marketable product. We present an example of a standard product market model (for determining prices and production quantities for two competing products in the aftermath of the technology competition) in which the assumptions of our technology competition model hold. An additional key finding is that there can be discontinuities in the returns that accrue from enhanced cost competence. If a firm is on the right side of a “jump point,” its expected profits can be dramatically more than if it were only slightly less competent. These jump points arise where the firm's cost competence becomes sufficient to cause a competitor to decide against competing in the new technology, or to significantly drop its investment amount. When a potential competitor backs down, the prospect of high returns for the incumbent opens up.

Delivering high quality service during the service encounter is central to competitive advantage in service organizations. However, achieving such high quality while controlling for costs is a major challenge for service managers. The purpose of this paper is to present an approach for addressing this challenge. The approach entails developing a model linking service process operational variables to service quality metrics to provide guidelines for service resource allocation. The approach enables the service operations manager to take specific actions toward service quality improvement, in light of the costs involved. A novel feature of the approach is the development of robust optimization models, which provide optimal operational guidelines while accounting for uncertainty in the model's parameters. We demonstrate the applicability of the approach in a large health care facility.

Suppose you are a Marketing Manager envisioning a new product, or an Operations Manager contemplating a process improvement, or a CEO who commissioned an integrated new product development team. If our assumptions hold, our model offers you a single numerical measure, called the degree of product/process innovation, to determine your initiative's impact on potential sales, prices, market segments, and profits. Our simple, single-period model is a variation of the existing vertically differentiated products model: There are two competing substitute products, and customers will buy at most one of them. Our contribution is to allow new relationships between the valuations of the two products by potential customers, and to allow differing unit production costs. We identify equilibrium results when two competing firms each offer one product, and find the profit maximizing result when one (monopolistic) firm offers both products. The new product infringes on the market in one of two ways: High-end encroachment results when the new product attracts the best customers (those with the highest reservation prices), while low-end encroachment identifies a situation where the new product attracts fringe (lower-end) customers. Low-end encroachment may help explain why an incumbent sometimes fails to recognize the threat of an entrant's product, as we illustrate with an example from the disk drive industry. In short, we offer insight into the value of both a marketing objective (enhancing the product design attributes) and a manufacturing goal (lowering the production cost) in a product and/or process improvement project.