Explore the vast range of titles available.

Given uncertain popularity of new products by location, fast fashion retailer Zara faces a trade-off. Large initial shipments to stores reduce lost sales in the critical first days of the product life cycle, but maintaining stock at the warehouse allows restocking flexibility once initial sales are observed. In collaboration with Zara, we develop and test a decision support system featuring a data-driven model of forecast updating and a dynamic optimization formulation for allocating limited stock by location over time. A controlled field experiment run worldwide with 34 articles during the 2012 season showed an increase in total average season sales by approximately 2% and a reduction in the number of unsold units at the end of the regular selling season by approximately 4%.

Thompson sampling is a randomized Bayesian machine learning method, whose original motivation was to sequentially evaluate treatments in clinical trials. In recent years, this method has drawn wide attention, as Internet companies have successfully implemented it for online ad display. In “Online network revenue management using Thompson sampling,” K. Ferreira, D. Simchi-Levi, and H. Wang propose using Thompson sampling for a revenue management problem where the demand function is unknown. A main challenge to adopt Thompson sampling for revenue management is that the original method does not incorporate inventory constraints. However, the authors show that Thompson sampling can be naturally combined with a linear program formulation to include inventory constraints. The result is a dynamic pricing algorithm that incorporates domain knowledge and has strong theoretical performance guarantees as well as promising numerical performance results. Interestingly, the authors demonstrate that Thompson sampling achieves poor performance when it does not take into account domain knowledge. Finally, the proposed dynamic pricing algorithm is highly flexible and is applicable in a range of industries, from airlines and internet advertising all the way to online retailing. We consider a price-based network revenue management problem in which a retailer aims to maximize revenue from multiple products with limited inventory over a finite selling season. As is common in practice, we assume the demand function contains unknown parameters that must be learned from sales data. In the presence of these unknown demand parameters, the retailer faces a trade-off commonly referred to as the “exploration-exploitation trade-off.” Toward the beginning of the selling season, the retailer may offer several different prices to try to learn demand at each price (“exploration” objective). Over time, the retailer can use this knowledge to set a price that maximizes revenue throughout the remainder of the selling season (“exploitation” objective). We propose a class of dynamic pricing algorithms that builds on the simple, yet powerful, machine learning technique known as “Thompson sampling” to address the challenge of balancing the exploration-exploitation trade-off under the presence of inventory constraints. Our algorithms have both strong theoretical performance guarantees and promising numerical performance results when compared with other algorithms developed for similar settings. Moreover, we show how our algorithms can be extended for use in general multiarmed bandit problems with resource constraints as well as in applications in other revenue management settings and beyond. The online appendix is available at https://doi.org/10.1287/opre.2018.1755 .

Because of uncertainty in customer demand and lack of understanding in customer reactions to price changes, it is a challenge for many companies, such as manufacturers and retailers, to match supply and demand. Most of the models in the operations literature, however, have focused on the case in which the underlying customer demand information is known a priori, which is not true in many applications. In “Coordinating Pricing and Inventory Replenishment with Nonparametric Demand Learning,” B. Chen, X. Chao, and H. Ahn develop a data-driven algorithm for pricing and inventory decisions that learns the demand and customer information from sales data on the fly, and they show that the profit generated from the algorithm converges to the clairvoyant optimal profit at the quickest possible rate. We consider a firm (e.g., retailer) selling a single nonperishable product over a finite-period planning horizon. Demand in each period is stochastic and price sensitive, and unsatisfied demands are backlogged. At the beginning of each period, the firm determines its selling price and inventory replenishment quantity with the objective of maximizing total profit, but it knows neither the average demand (as a function of price) nor the distribution of demand uncertainty a priori; hence, it has to make pricing and ordering decisions based on observed demand data. We propose a nonparametric, data-driven algorithm that learns about the demand on the fly and, concurrently, applies learned information to make replenishment and pricing decisions. The algorithm integrates learning and action in a sense that the firm actively experiments on pricing and inventory levels to collect demand information with minimum profit loss. Besides convergence of optimal policies, we show that the regret of the algorithm, defined as the average profit loss compared with that of the optimal solution had the firm known the underlying demand information, vanishes at the fastest possible rate as the planning horizon increases.

This paper studies the potential benefits of responsive pricing and demand learning to sellers of seasonal fashion goods. As typical in such markets, demand uncertainty is high at the beginning of a season, but there is a potential opportunity to learn about demand via early sales observations. Additionally, although the consumers have general preference for purchasing a fashion product earlier rather than later in the season, they may exhibit strategic behavior—contemplating the benefits of postponing their purchase in anticipation of end-of-season discounts. Our results demonstrate that the benefits of responsive pricing, in comparison with a benchmark case of a fixed-price policy, depend sharply on the nature of the consumers’ behavior. Interestingly, in stark contrast to markets of myopic consumers, when the consumers are all strategic, the benefits of responsive pricing tend to worsen when there is a higher potential for learning. We explain this counterintuitive outcome by pointing to two phenomena: the spread effect and information shaping. For example, sellers of fashion products that consider upgrading their pricing systems to incorporate “ accurate response ” strategies (i.e., integrating learning and responsive pricing) should be aware of the possibility that such action might lead them to a new and potentially worse equilibrium, particularly when there is a higher opportunity to learn. Despite the fact that price commitment completely eliminates the seller’s ability to learn, it appears to increasingly dominate responsive pricing as the portion of strategic consumers in the market increases. But, although performing better than responsive pricing, a price-commitment policy is typically limited in performing effective discrimination. Finally, we studied the potential benefits of quick response strategies—ones that embed both dynamic pricing and quick inventory replenishment during the sales season—and found that they are particularly significant under strategic consumer behavior. We explain this result by arguing that quick response provides the seller with a real option that serves as an effective implicit threat to the consumers: encouraging them to buy earlier at premium prices rather than wait for discounts at the end of the season. The online appendix is available at https://doi.org/10.1287/mnsc.2018.3114 . This paper was accepted by Martin Lariviere, operations management.

In the conceptualization of thriving at work, it is emphasized that employees’ learning and vitality are two equally important components of thriving and that thriving is facilitated by contextual features and available resources. In this study, we examined the effects of two challenge stressors (time pressure and learning demands) on thriving at work. Based on the literature on challenge and hindrance stressors, we proposed that challenge stressors positively affect learning and negatively affect vitality. To uncover underlying mechanisms, we measured challenge appraisal and hindrance appraisal of work situations in a diary study. A sample of 124 knowledge workers responded to three daily surveys (before the lunch break, during the afternoon, and at the end of the workday) for a period of five workdays. Results indicate that the indirect effects of learning demands and time pressure on learning are mediated by challenge appraisal, whereas indirect effects of learning demands on vitality are mediated by hindrance appraisal. Overall, our study shows that challenge stressors have a positive total effect on learning but no total effect on vitality. These differential relationships call for a finer distinction between the two components of thriving at work in future research.

We consider an inventory control problem with multiple products and stockout substitution. The firm knows neither the primary demand distribution for each product nor the customers’ substitution probabilities between products a priori, and it needs to learn such information from sales data on the fly. One challenge in this problem is that the firm cannot distinguish between primary demand and substitution (overflow) demand from the sales data of any product, and lost sales are not observable. To circumvent these difficulties, we construct learning stages with each stage consisting of a cyclic exploration scheme and a benchmark exploration interval. The benchmark interval allows us to isolate the primary demand information from the sales data, and then this information is used against the sales data from the cyclic exploration intervals to estimate substitution probabilities. Because raising the inventory level helps obtain primary demand information but hinders substitution demand information, inventory decisions have to be carefully balanced to learn them together. We show that our learning algorithm admits a worst-case regret rate that (almost) matches the theoretical lower bound, and numerical experiments demonstrate that the algorithm performs very well. This paper was accepted by J. George Shanthikumar, big data analytics.

“Bespoke,” or mass customization strategy, combines demand learning and preference learning. We develop an analytical framework to study the economic value of bespoke systems and investigate the interaction between demand learning and preference learning. We find that it is possible for demand learning and preference learning to be either complements or substitutes, depending on the customization cost and the demand uncertainty profile. They are generally complements when the personalization cost is low and the probability of having high demand is large. Contrary to usual belief, we show that higher demand uncertainty does not necessarily yield more complementarity benefits. Our numerical study shows that the complementarity benefit becomes weaker when customers are more strategic. Interestingly, the substitute loss can occur when the personalization cost is small and the probability of having high demand is large, when customers are strategic. The online supplement is available at https://doi.org/10.1287/mnsc.2017.2771 . This paper was accepted by Serguei Netessine, operations management.

\"Bespoke,\" or mass customization strategy, combines demand learning and preference learning. We develop an analytical framework to study the economic value of bespoke systems and investigate the interaction between demand learning and preference learning. We find that it is possible for demand learning and preference learning to be either complements or substitutes, depending on the customization cost and the demand uncertainty profile. They are generally complements when the personalization cost is low and the probability of having high demand is large. Contrary to usual belief, we show that higher demand uncertainty does not necessarily yield more complementarity benefits. Our numerical study shows that the complementarity benefit becomes weaker when customers are more strategic. Interestingly, the substitute loss can occur when the personalization cost is small and the probability of having high demand is large, when customers are strategic.

PurposeThis study aimed at determining factors which affect university lecturers’ adoption of the Moodle platform under the conditions of COVID-19. In considering the condition of the COVID-19 pandemic, the unified theory of acceptance and use of technology (UTAUT) model was applied and extended by adding two additional variables of learning demand and time pressure to assess their influence on Moodle platform adoption.Design/methodology/approachData were obtained from the 226 participants through an online structured questionnaire. The covariance-based approach of structural equation modeling was used to examine the proposed model. The structural model was tested using the maximum likelihood method of analysis of a moment structures to analyze the study’s hypotheses.FindingsResults suggest that performance expectations have a substantial influence on behavioral intent. The effort expectancy, social effect and facilitative factors have no effects on behavioral intentions. Facilitating conditions directly and significantly affect the actual use of Moodle. The results also reveal that learning demands, which is a salient predictor of perceived time pressure, in turn directly and significantly affects the actual use of Moodle. Finally, the behavioral intention has a strong influence on Moodle’s actual usage.Originality/valueAlthough the UTAUT 2 model is considered to be a new and updated version of UTAUT, it has not been used since newly added variables, namely, price, habit and hedonic motivations, are less related to the context and to avoid respondents’ paradox. Moreover, using the Moodle platform in the researched context is compulsory for both students and instructors. Discussion, insights, limitations and recommendations for future studies are suggested.

The price demand relation is a fundamental concept that models how price affects the sale of a product. It is critical to have an accurate estimate of its parameters, as it will impact the company’s revenue. The learning has to be performed very efficiently using a small window of a few test points, because of the rapid changes in price demand parameters due to seasonality and fluctuations. However, there are conflicting goals when seeking the two objectives of revenue maximization and demand learning, known as the learn/earn trade-off. This is akin to the exploration/exploitation trade-off that we encounter in machine learning and optimization algorithms. In this paper, we consider the problem of price demand function estimation, taking into account its exploration–exploitation characteristic. We design a new objective function that combines both aspects. This objective function is essentially the revenue minus a term that measures the error in parameter estimates. Recursive algorithms that optimize this objective function are derived. The proposed method outperforms other existing approaches.