Coordinating Service‐Sensitive Demand and Capacity by Adaptive Decision Making: An Application of Case‐Based Decision Theory*

The subject of this article is the simultaneous choice of product price and manufacturing capacity if demand is stochastic and service‐level sensitive. In this setting, capacity as well as price have an impact on demand because several aspects of service level depend on capacity. For example, delivery time will be reduced if capacity is increased given a constant demand rate. We illustrate the relationship between service level, capacity, and demand reaction by a stylized application problem from the after‐sales services industry. The reaction of customers to variations in service level and price is represented by a kinked price‐demand‐rate function. We first derive the optimal price‐capacity combination for the resulting decision problem under full information. Subsequently, we focus on a decision maker (DM) who lacks complete knowledge of the demand function. Hence the DM is unable to anticipate the service level and consequently cannot identify the optimal solution. However, the DM will acquire additional information during the sales process and use it in subsequent revisions of the price‐capacity decision. Thus, this decision making is adaptive and based on experience. In contrast to the literature, which assumes certain repetitive procedures somewhat ad hoc, we develop an adaptive decision process based on case‐based decision theory (CBDT) for the price‐capacity problem. Finally, we show that a CBDT DM in our setting eventually finds the optimal solution, if the DM sets the price based on absorption costs and adequately adjusts the capacity with respect to the observed demand.     

Pareto‐Improving Contracts for Express Package Delivery Services

We address the problem of an express package delivery company in structuring a long‐term customer contract whose terms may include prices that differ by day‐of‐week and by speed‐of‐service. The company traditionally offered speed‐of‐service pricing to its customers, but without day‐of‐week differentiation, resulting in customer demands with considerable day‐of‐week seasonality. The package delivery company hoped that using day‐of‐week and speed‐of‐service price differentiation for contract customers would induce these customers to adjust their demands to become counter‐cyclical to the non‐contract demand. Although this usually cannot be achieved by pricing alone, we devise an approach that utilizes day‐of‐week and speed‐of‐service pricing as an element of a Pareto‐improving contract. The contract provides the lowest‐cost arrangement for the package delivery company while ensuring that the customer is at least as well off as he would have been under the existing pricing structure. The contract pricing smoothes the package delivery company's demand and reduces peak requirements for transport capacity. The latter helps to decrease capital costs, which may allow a further price reduction for the customer. We formulate the pricing problem as a biconvex optimization model, and present a methodology for designing the contract and numerical examples that illustrate the achievable savings.     

Lead Time and Price Quotation Mode Selection: Uniform or Differentiated?

Firms in service and make‐to‐order manufacturing industries often quote lead times and prices to customers. We define uniform quotation mode (UQM) as the strategy where a firm offers a single lead time and price quotation, and differentiated quotation mode (DQM) is where a firm offers a menu of lead times and prices for customers to choose from. Both modes are followed in practice. Firms should determine which is more profitable. We classify customers into two groups: lead time sensitive (LS) and price sensitive (PS). LS customers value lead time reduction more than PS customers. We develop mathematical models of both quotation modes and analyze them to determine the most profitable mode under specified situations as well as the best lead time and price quotations within each mode. We find that DQM is dominated by UQM whenever PS customers have positive utilities from UQM or LS customers have positive utilities from DQM. Otherwise, which quotation mode is better depends on multiple factors, such as customer characteristics (including lead time reduction valuation and product valuation of a customer, and the proportion of LS customers) and production characteristics (including the desired service level and service or production cost).     

Service Systems with Experience‐Based Anecdotal Reasoning Customers

The existing queueing literature typically assumes that customers either perfectly know the expected waiting time or are able to form rational expectations about it. In contrast, in this article, we study canonical service models where customers do not have such full information or capability. We assume that customers lack full capability or ample opportunities to perfectly infer the service rate or estimate the expected waiting time, and thus can only rely on past experiences and anecdotal reasoning to make their joining decisions. We fully characterize the steady‐state equilibrium in this service system. Compared with the fully rational benchmark, we find that customers with anecdotal reasoning are less price‐sensitive. Consequently, with a higher market potential (higher arrival rate), a revenue‐maximizing firm may increase the price if the service rate is exogenous, and it may decrease the price if the service rate is at the firm's discretion. Both results go against the commonly accepted pricing recommendations in the fully rational benchmark. We also show that revenue maximization and welfare maximization lead to fundamentally different pricing strategies with anecdotal reasoning, whereas they are equivalent in the fully rational benchmark.     

Capacity rationing in rental systems with two customer classes and batch arrivals

Service differentiation is an emerging method to improve profit and to better serve high-priority customers. Such an approach has recently been introduced by one of Europe's leading rail cargo companies. Under this approach, customers can choose between classic and premium services. Premium service is priced above classic service and premium customers receive a service guarantee which classic customers do not receive. The company has to decide under which conditions it should ration its fleet capacity to classic customers in order to increase service of premium customers. We model such a situation as a batch-arrival queuing loss system. We describe the model, solve it optimally, and derive quantities of interest such as service probabilities. We further analyze it by performing numerical experiments based on the data from the company that motivated our research. We show that the potential of capacity rationing can be substantial in situations like the one we analyzed. We also derive conditions under which rationing is especially beneficial, such as under high unit fleet holding costs or in the presence of batch arrivals compared to single arrivals.     

Self‐Service Operations at Retail Stores: The Role of Inter‐Customer Interactions

Inter‐customer interactions are important to the operation of self‐services in retail settings. More specifically, when self‐service terminals are used as part of customers’ checkout processes in retail operations without the explicit involvement of retailers as the direct service providers, inter‐customer interactions become a significant managerial issue. In this article, we examine the impact of inter‐customer interactions at retail self‐service terminals on customers’ service quality perceptions and repeat purchase intentions at retail stores. We conduct a scenario‐based experimental design (N = 674) using a 2 × 2 factorial design in which inter‐customer interactions are divided into “positive” vs. “negative” and occur during the “waiting” or during the actual “transaction” stages of self‐services at a retail store. We use attribution theory to develop the hypotheses. The results demonstrate that, through their interactions, fellow customers can exert influences on a focal customer's quality perceptions and repeat purchasing intentions toward a retail store. Furthermore, these influences were impacted by how customers attribute blame or assign responsibility toward the retail store. Service operations managers should leverage these interactions by designing into self‐service settings the capacities and interfaces that are best suited for customers’ co‐production of their self‐service experiences.     

Self‐Selecting Priority Queues with Burr Distributed Waiting Costs

Service providers, in the presence of congestion and heterogeneity of customer waiting costs, often introduce a fee‐based premier option using which the customers self‐segment themselves. Examples of this practice are found in health care, amusement parks, government (consular services), and transportation. Using a single‐server queuing system with customer waiting costs modeled as a Burr Distribution, we perform a detailed analysis to (i) determine the conditions (fees, cost structure, etc.) under which this strategy is profitable for the service provider, (ii) quantify the benefits accrued by the premier customers; and (iii) evaluate the resulting impact on the other customers. We show that such self‐selecting priority systems can be pareto‐improving in the sense that they are beneficial to everyone. These benefits are larger when the variance in the customer waiting costs is high and the system utilization is high. We use income data from the poorest and richest areas (identified by zipcode) in the United States along with the countrywide income distribution to illustrate our results. Numerical results indicate that planning for a 20–40% enrollment in the high‐priority option is robust in ensuring that all the stakeholders benefit from the proposed strategy.     

Impeding the Juggernaut of Innovation Diffusion: A Production‐Constrained Model

Models of innovation diffusion typically depict an inexorable momentum once the process begins to roll. Limited production capacity, however, can place a cap on this process, leading to waiting lines of potential customers, thus diminishing overall service quality and the speed of diffusion. Identifying the minimum production capacity needed for unimpeded and unimpaired diffusion can ensure that there are no customers waiting to adopt the product. We propose a production‐capacity‐constrained diffusion model that considers an exogenous industry production capacity and accounts for word‐of‐mouth effects from adopters as well as waiting customers. We derive analytical expressions for minimum capacity needed under multiple production scenarios. We present a dual‐objective non‐linear least squares procedure with large‐scale grid search for estimating the parameters. We apply our model to several new product innovation data sets, ranging from vacuum cleaners to sports utility vehicles in the United States to iPhones globally. Our estimates show that product shortages exist, ranging from mild to severe, in all of these product markets. We are able to corroborate some of our findings with independent external sources of evidence. We find that information on industry capacity can be recovered with as few as 5 years of sales data. Our model has practical implications for policy makers and can help equity analysts triangulate industry capacity better, particularly when such information is closely held.     

COMPONENT COMMONALITY: MODELS WITH PRODUCT‐SPECIFIC SERVICE CONSTRAINTS*

Common components are used extensively for reasons including product postponement and expediting new product development. We consider a two‐stage assemble‐to‐order system with two products having uniformly distributed demand, one common component, and product‐specific components. We develop optimization models in which the cost‐minimizing inventory of the components must be determined and allocated to products in order to meet product‐specific service level constraints. We compare two different commonality models based on whether or not the products are prioritized. A distinctive feature of our study is the use of product‐specific service levels. We compare our results with models using aggregate service levels.     

Skill Management in Large‐Scale Service Marketplaces

Large‐scale, web‐based service marketplaces have recently emerged as a new resource for customers who need quick resolutions for their short‐term problems. Due to the temporary nature of the relations between customers and service providers (agents) in these marketplaces, customers may not have an opportunity to assess the ability of an agent before their service completion. On the other hand, the moderating firm has a more sustained relationship with agents, and thus it can provide customers with more information about the abilities of agents through skill screening mechanisms. In this study, we consider a marketplace where the moderating firm can run two skills tests on agents to assess if their skills are above certain thresholds. Our main objective is to evaluate the effectiveness of skill screening as a revenue maximization tool. We, specifically, analyze how much benefit the firm obtains after each additional skill test. We find that skill screening leads to negligible revenue improvements in marketplaces where agent skills are highly compatible and the average service times are similar for all customers. As the compatibility of agent skills weakens or the customers start to vary in their processing time needs, we show that the firm starts to experience sizable improvements in revenue from skill screening. Apparently, the firm can reap the most of these substantial benefits when it runs only one test. For instance, in marketplaces where agents posses uncorrelated skills, the second skill test only brings an additional 2% improvement in revenue. Accounting for possible skill screening costs, we then show the optimality of offering only one test when the compatibility between agent skills is sufficiently low. The results of this study also have important implications in terms of the right level of intervention in the marketplaces we study.     

TIME‐BASED PRICING AND LEADTIME POLICIES FOR A BUILD‐TO‐ORDER MANUFACTURER

We studied time‐based policies on pricing and leadtime for a build‐to‐order and direct sales manufacturer. It is assumed that the utility of the product varies among potential customers and decreases over time, and that a potential customer will place an order if his or her utility is higher than the manufacturer's posted price. Once an order is placed, it will be delivered to the customer after a length of time called “leadtime.” Because of the decrease in a customer's utility during leadtime, a customer will cancel the order if the utility falls below the ordering price before the order is received. The manufacturer may choose to offer discounted prices to customers who would otherwise cancel their orders. We discuss two price policies: common discounted price and customized discounted price. In the common discounted price policy, the manufacturer offers a single lower price to the customers; in the customized discounted price policy, the manufacturer offers the customers separately for individual new prices. Our analytical and numerical studies show that the discounted price policies results in higher revenue and that the customized discounted price policy significantly outperforms the common discounted price policy when product utility decreases rapidly. We also study two leadtime policies when production cost decreases over time. The first uses a fixed leadtime, and the second allows the leadtime to vary dynamically over time. We find that the dynamic leadtime policy significantly outperforms the fixed leadtime policy when the product cost decreases rapidly.     

Using Frontier Portfolios to Improve Make‐to‐Order Operations

Make‐to‐order (MTO) manufacturers face a common problem of maintaining a desired service level for delivery at a reasonable cost while dealing with irregular customer orders. This research considers a MTO manufacturer who produces a product consisting of several custom parts to be ordered from multiple suppliers. We develop procedures to allocate orders to each supplier for each custom part and calculate the associated replenishment cost as well as the probability of meeting the delivery date, based on the suppliers' jobs on hand, availability, process speed, and defective rate. For a given delivery due date, a frontier of service level and a replenishment cost frontier are created to provide a range of options to meet customer requirements. This method can be further extended to the case when the delivery due date is not fixed and the manufacturer must “crash” its delivery time to compete for customers.     

A Stochastic Inventory Model With Fast‐Ship Commitments

We present a multiperiod model of a retail supply chain, consisting of a single supplier and a single retailer, in which regular replenishment occurs periodically but players have the option to support fast delivery when customers experience a stockout during a replenishment period. Because expedited shipments increase the supplier's transportation cost, and possibly production/inventory costs, the supplier typically charges a markup over and above the prevailing wholesale price for fast‐shipped items. When fast shipping is not supported, items are backordered if customers are willing to wait until the start of the next replenishment period. We characterize the retailers and the supplier's optimal stocking and production policies and then utilize our analytical framework to study how the two players respond to changes in supply chain parameters. We identify a sufficient condition such that the centralized supply chain is better off with the fast‐ship option. We find a range of markups for fast‐ship orders such that the fast‐ship option is preferred by both the supplier and the retailer in a decentralized chain. However, a markup that is a win–win for both players may not exist even when offering fast‐ship option is better for the centralized chain. Our analysis also shows that depending on how the markup is determined, greater customer participation in fast‐ship orders does not necessarily imply more profits for the two players. For some predetermined markups, the retailer's profit with the fast‐ship option is higher when more customers are willing to wait. However, the retailer may not be able to benefit from the fast‐ship option because the supplier may choose not to support the fast‐ship option when fast‐ship participation increases due to the fact that the fast‐ship participation rate adversely affects the initial order size.     

Measuring Seat Value in Stadiums and Theaters

We study how the seat value perceived by consumers attending an event in a theater/stadium depends on the location of their seat relative to the stage/field. We develop a measure of seat value, called the Seat Value Index, and relate it to seat location and consumer characteristics. We implement our analysis on a proprietary data set that a professional baseball franchise in Japan collected from its customers, and provide recommendations. For instance, we find that customers seated in symmetric seats on left and right fields might derive very different valuations from the seats. We also find that the more frequent visitors to the stadium report extreme seat value less often when compared with first‐time visitors. Our findings and insights remain robust to the effects of price and game‐related factors. Thus, our research quantifies the significant influence of seat location on the ex‐post seat value perceived by customers. Utilizing the heterogeneity in seat values at different seat locations, we provide segment‐specific pricing recommendations based on a service‐level objective that would limit the fraction of customers experiencing low seat value to a desired threshold.     

Opt‐Out Options in New Product Co‐development Partnerships

We study three contractual arrangements—co‐development, licensing, and co‐development with opt‐out options—for the joint development of new products between a small and financially constrained innovator firm and a large technology company, as in the case of a biotech innovator and a major pharma company. We formulate our arguments in the context of a two‐stage model, characterized by technical risk and stochastically changing cost and revenue projections. The model captures the main disadvantages of traditional co‐development and licensing arrangements: in co‐development the small firm runs a risk of running out of capital as future costs rise, while licensing for milestone and royalty (M&R) payments, which eliminates the latter risk, introduces inefficiency, as profitable projects might be abandoned. Counter to intuition we show that the biotech's payoff in a licensing contract is not monotonically increasing in the M&R terms. We also show that an option clause in a co‐development contract that gives the small firm the right but not the obligation to opt out of co‐development and into a pre‐agreed licensing arrangement avoids the problems associated with fully committed co‐development or licensing: the probability that the small firm will run out of capital is greatly reduced or completely eliminated and profitable projects are never abandoned.     

An Advanced Heuristic for Multiple‐Option Spare Parts Procurement after End‐of‐Production

After‐sales service is a major source of profit for many original equipment manufacturers in industries with durable products. Successful engagement in after‐sales service improves customer loyalty and allows for competitive differentiation through superior service like an extended service period during which customers are guaranteed to be provided with service parts. Inventory management during this period is challenging due to the substantial uncertainty concerning demand over a long time horizon. The traditional mechanism of spare parts acquisition is to place a large final order at the end of regular production of the parent product, causing major holding costs and a high level of obsolescence risk. With an increasing length of the service period, more flexibility is needed and can be provided by adding options like extra production and remanufacturing. However, coordinating all three options yields a complicated stochastic dynamic decision problem. For that problem type, we show that a quite simple decision rule with order‐up‐to levels for extra production and remanufacturing is very effective. We propose a heuristic procedure for parameter determination which accounts for the main stochastic and dynamic interactions in decision making, but still consists of relatively simple calculations that can be applied to practical problem sizes. A numerical study reveals that the heuristic performs extremely well under a wide range of conditions, and therefore can be strongly recommended as a decision support tool for the multi‐option spare parts procurement problem. A comparison with decision rules adapted from practice demonstrates that our approach offers an opportunity for major cost reductions.     

Electricity Time‐of‐Use Tariff with Stochastic Demand

In this article, we study the electricity time‐of‐use (TOU) tariff for an electricity company with stochastic demand. The electricity company offers the flat rate (FR) and TOU tariffs to customers. Under the FR tariff, the customer pays a flat price for electricity consumption in both the peak and non‐peak periods. Under the TOU tariff, the customer pays a high price for electricity consumption in the peak period and a low price for electricity consumption in the non‐peak period. The electricity company uses two technologies, namely the base‐load and peak‐load technologies, to generate electricity. We derive the optimal capacity investment and pricing decisions for the electricity company. Furthermore, we use real data from a case study to validate the results and derive insights for implementing the TOU tariff. We show that in almost all the cases, the electricity company needs less capacity for both technologies under the TOU tariff than under the FR tariff, even though the expected demand in the non‐peak period increases. In addition, except for some extreme cases, there is essentially no signicant reduction in the total demand of the two periods, although the TOU tariff can reduce the demand in the peak period. Under the price‐cap regulation, the customer may pay a lower price on average under the TOU tariff than under the FR tariff. We conduct an extensive numerical study to assess the impacts of the model parameters on the optimal solutions and the robustness of the analytical results, and generate managerial implications of the research findings.     

Lot‐sizing Decisions Under Limited‐time Price Reduction*

We consider the optimal lot‐sizing policy for an inventoried item when the vendor offers a limited‐time price reduction. We use the discounted cash flow (DCF) approach in our analysis, thereby eliminating the sources of approximation found in most of the earlier studies that use an average annual cost approach. We first characterize the optimal lot‐sizing policies and their properties, then develop an algorithm for determining the optimal lot sizes. We analytically demonstrate that the lot sizes derived using an average annual cost approach for the different variants of the problem are, in general, larger than the DCF optimum. While DCF analysis is more rigorous and yields precise lot sizes, we recognize that the associated mathematical models and the solution procedure are rather complex. Since simple and easy‐to‐understand policies have a strong practical appeal to decision makers, we propose a DCF version of a simple and easy‐to‐implement heuristic called the “Early Purchase” (EP) strategy and discuss its performance. We supplement our analytical developments with a detailed computational analysis and discuss the implications of our findings for decision making.     

Equilibrium Joining Strategies and Optimal Control of a Make‐to‐Stock Queue

We consider a make‐to‐stock, finite‐capacity production system with setup cost and delay‐sensitive customers. To balance the setup and inventory related costs, the production manager adopts a two‐critical‐number control policy, where the production starts when the number of waiting customers reaches a certain level and shuts down when a certain quantity of inventory has accumulated. Once the production is set up, the unit production time follows an exponential distribution. Potential customers arrive according to a Poisson process. Customers are strategic, i.e., they make decisions on whether to stay for the product or to leave without purchase based on their utility values, which depend on the production manager's control decisions. We formulate the problem as a Stackelberg game between the production manager and the customers, where the former is the game leader. We first derive the equilibrium customer purchasing strategy and system performance. We then formulate the expected cost rate function for the production system and present a search algorithm for obtaining the optimal values of the two control variables. We further analyze the characteristics of the optimal solution numerically and compare them with the situation where the customers are non‐strategic.     

Dynamic Pricing in e‐Services under Demand Uncertainty

High volatility of the e‐services market, due to increasing competition, low life cycle of products, and easy availability of information about competing service offerings to customers, makes the demand for service offerings quite uncertain. Revenue management in such markets calls for real‐time techniques to learn the demand and its dependence on both the price and the service level associated with the service offering. We assume firms reply on exploratory approaches for demand estimation, in which firms experiment with different service offerings in order to simultaneously learn the demand while doing business. Such exploration and learning process can be costly without supervision. As reported by Rothschild (Journal of Economic Theory, 9 185‐202, 1974), traditional Bayesian dynamic control approaches may conclude with suboptimal offerings. We propose a novel demand learning approach that is guaranteed to converge to the optimal offering. The approach combines simulated annealing algorithm with Bayesian learning. We further present intelligent techniques that adaptively reduce the effort of exploration on suboptimal service offerings so as to improve the long‐run average profit.