Knowledge that Transforms

Omni-channel retailers typically face high return rates, particularly in their online channel. This paper examines how these returns can be converted into exchanges by leveraging omni-channel capabilities. We do so in the context of a fast-fashion retailer running integrated physical store and online channels. Our unit of analysis is a return journey, which starts with an initial online purchase, proceeds with a return, and then potentially continues with subsequent exchanges and returns, replacing the original purchase. Using a random forest with instrumental variables, we study whether store visits for pick-up or return of the initial purchase influence a consumer’s likelihood to make an exchange, thereby generating revenue for the retailer. We explore the heterogeneous effect of these store visits across different customer profiles based on the recency-frequency-monetary value (RFM) framework. Our results indicate that store visits for pick-up or return increase the likelihood of an exchange, with a more pronounced effect among less valuable customers. Follow-up analyses suggest that this effect is driven by the customers’ ability to reduce product uncertainty through in-store inspection. Additionally, a store visit decreases the likelihood of a return (consistent with previous literature) and increases the likelihood of keeping an exchange purchase when it has been made. Our findings thus underscore the critical role of physical stores for online shoppers in finding the right product and, consequently, for retailers to convert returns into exchanges.

Armed conflicts, terrorism, and political instability disrupt the implementation of thousands of prosocial projects in the developing world. In this unstable, fast-paced environment, managers may need to change course and adapt projects to new circumstances. Could adaptive management be a solution? Leveraging a groundbreaking pilot run by the World Bank, we evaluate the performance of all 429 adaptive projects implemented between 1998 and 2013, when the Bank launched the “adaptable program loan.” Our study—covering 91 countries, 25 development sectors, and approximately $55 billion in funding resources—examines how armed conflict and adaptability influence project performance. We find that adaptability, reinforced by experience and learning, counteracts conflict’s negative effect on project performance. However, this moderating effect diminishes as project complexity increases and reverses in non-conflict areas due to cost overruns and overoptimistic formulations .

The swift advancement of social technologies has created unprecedented opportunities for companies to embrace new business models or bolster their existing ones. However, the adoption of controversial technologies with social characteristics can also expose firms to risks, as these technologies may not be entirely under their operational control. In this study, our objective is to analyze the impact of the adoption of a controversial technology (i.e., cryptocurrency payment) on users’ perception of a firm and firm performance and examine how social media influences this impact. We leverage a unique research context where an online travel agency cooperates with Travala.com for cryptocurrency payment. Using conventional and synthetic difference-in-differences and randomized experiments, we compare the general perceptions of customers towards the online travel agency before and after the cooperation and gain three main findings. First, the adoption of cryptocurrency payment generally leads to a significant decrease in revenue because of the negative associations concerning cryptocurrency. Second, the popularity of cryptocurrency payment on social media mitigates this negative effect. Third, when cryptocurrency prices rise and public opinions on cryptocurrency turn positive, the mitigation effect is more salient. Further analysis reveals that our results are robust after controlling for the seasonality effect and the nature of the collaboration and using different lags for the outcome variable. This study is among the first to empirically examine how cryptocurrency payment adoption affects firm performance, and provide important theoretical and practical implications regarding the adoption of controversial technologies and the externalities of social media.

We consider a multilocation newsvendor network in which historical data are the only available information about the joint demand distribution. To determine optimal inventory levels, we develop a novel data-driven two-stage distributionally robust optimization model that does not assume the demand support is known. Instead, we infer the support from historical data using two prediction algorithms, which yield quantile-based and Mahalanobis-distance-based support estimates and therefore either ignore or capture cross-location demand dependence. Our objective is to minimize worst-case expected cost over an ambiguity set constructed from these support estimates, consisting of all probability distributions within a prescribed type- <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:mi mathvariant="normal">∞</mml:mi> </mml:math> Wasserstein ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:msub> <mml:mrow> <mml:mi mathvariant="sans-serif">W</mml:mi> </mml:mrow> <mml:mi mathvariant="normal">∞</mml:mi> </mml:msub> </mml:math> ) distance of the empirical distribution. To approximate the second-stage recourse decisions, we employ a multiple-linear-decision-rule approximation that is provably asymptotically optimal. This leads to tractable linear programming and second-order cone programming reformulations for the quantile-based and Mahalanobis-distance-based formulations, respectively. We also establish support-aware finite-sample guarantees for the proposed framework. Numerical results show that quantile-based support estimation is more effective at maintaining reliable service levels, whereas Mahalanobis-distance-based support estimation yields larger cost reductions, particularly under correlated demand.

The use of urgent care centers (UCCs) to reduce emergency department (ED) overuse has yielded mixed results in prior research. This paper investigates how the spatial proximity between UCCs and EDs drives heterogeneous effects of UCCs on ED demand. Using novel, granular encounter-level data from both EDs and newly opened UCCs within a major medical system, we leverage exogenous variation driven by institutional UCC operating hour policies to estimate causal effects. Through difference-in-differences models and generalized synthetic control methods, we find that collocated UCCs reduce ED demand for initial patient visits during UCC operating hours, with reductions spanning both nonurgent and urgent encounters. In contrast, non-collocated UCCs do not reduce ED demand. Such a difference is statistically significant. We identify queue sampling as the key mechanism driving these heterogeneous effects. Using timestamped arrival data, we show that higher ED waiting room census leads to increased arrivals at the collocated UCC, but only when the waiting room reaches visibly congested levels—in our setting, the top quintile of census distribution. This suggests patients exploit minimal switching costs to avoid observable congestion. This dynamic substitution pattern is absent at non-collocated sites, where transportation costs inhibit real-time facility switching. Our work sheds light on UCC location decisions and patients' facility choice behavior in an era of rapid UCC market expansion. For hospital managers, we show that when reducing ED demand is a primary objective, opening UCCs collocated with EDs proves substantially more effective than establishing them elsewhere in the region. For policymakers, while current UCC siting discussions focus primarily on expanding geographic access, our results demonstrate that carefully considering UCCs’ spatial configuration relative to EDs can unlock additional operational benefits beyond long-term access improvements.

This study examines manufacturers’ (i.e., brand-name firms’) incentives to adopt blockchain technology-supported (BTS) platforms in markets where deceptive and non-deceptive counterfeits coexist. It is among the first to analyze the adoption of BTS platforms in the context of a complex counterfeit market. First, the study integrates consumers’ perceived risk concerning product authenticity (CPRPA) into a signaling game model to examine the impacts of quality information asymmetry on market dynamics. The findings reveal that even though heightened CPRPA incentivizes deceptive counterfeiters to engage in non-deceptive sales (disclosing the true quality of their products), quality information asymmetry still exerts adverse impacts on both manufacturers and consumers, while simultaneously benefiting all counterfeiters. Second, the study evaluates manufacturers’ optimal strategies for adopting BTS platforms and the value derived from such adoption. The results suggest that CPRPA exerts a non-monotonic impact on manufacturers’ optimal decisions to adopt BTS platforms, and manufacturers may never benefit from such adoption in the presence of deceptive sales. Although higher adoption costs diminish the effectiveness of BTS platforms, such adoption can reduce illegal profits and improve consumer surplus. However, the adoption of BTS platforms may inadvertently increase demand for deceptive counterfeits. Consequently, BTS platforms are found to exert a more marked effect in combating non-deceptive counterfeits than deceptive ones. Additionally, BTS platform adoption contributes more to improving consumer surplus than to combating counterfeits.

Product returns are prevalent in practice. Many retailers provide lenient free return policies but with specific return window within which customers are allowed to return products. Motivated by this phenomenon, we consider a single-product online learning and pricing problem with stochastic product returns. A salient feature is that the demand function, depending on price and return window decisions, is initially unknown and must be learned on the fly. The retailer thus faces the classic exploration–exploitation trade-off. Moreover, we consider an inventory constraint, introducing an additional trade-off between earning revenue and managing inventory. We propose a modeling framework to integrate pricing and return window decisions, and develop a deterministic fluid model that serves as the full-information benchmark. To tackle the learning problem, we design a novel nonparametric learning algorithm that seamlessly integrates inverse stochastic gradient descent (SGD) and Upper Confidence Bound (UCB) methods. Under mild assumptions on demand and revenue functions, we establish a regret upper bound for our learning algorithm as <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:mi>O</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:msqrt> <mml:mi>W</mml:mi> <mml:mi>T</mml:mi> </mml:msqrt> <mml:mi>log</mml:mi> <mml:mspace width="0.2em"/> <mml:mi>T</mml:mi> <mml:mo stretchy="false">)</mml:mo> </mml:math> , where <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:mi>W</mml:mi> </mml:math> denotes the number of return window candidates and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:mi>T</mml:mi> </mml:math> denotes the time horizon. This result aligns with lower bounds established in both online pricing and multi-armed bandit (MAB) literature. Numerical experiments are conducted to verify the effectiveness and robustness of our algorithm across various environments. From an operational standpoint, retailers can use our learning framework as a decision-support tool to identify the optimal price and return window.

Platform owners employ strategic promotions (e.g., platform-initiated penetration pricing) to enlarge their customer base. Under the conventional price-driven economic logic, such promotions are expected to divert demand for the identical goods away from competing platforms, as extensively documented in the conventional brick-and-mortar store context. However, they can also boost the demand. In environments where customers can easily access the identical goods on competing platforms, the increased awareness by strategic promotion, coupled with other non-price factors, can stimulate the purchases of such goods on competing platforms. As such, the effect of strategic promotions on competitive dynamics remains unclear despite their importance for platforms and today’s intensifying platform competition. We utilize the context of online PC game marketplaces where Steam and Epic Games Store dominate. Using the synthetic control approach, our analysis finds that the strategic promotion by Epic Games Store, which offers selected games for free for a week, rather increases the sales of the identical games on Steam substantially (by 59.2%) during that week. In essence, although consumers could obtain the games for free on the promoting platform, more consumers purchased the identical games on the competing platform than before. To comprehensively understand the drivers and conditions of this phenomenon, we perform complementary analysis based on the Attention-Interest-Desire-Action framework, following the progression of consumers’ cross-platform purchase decision process. Using consumer survey and product review data, along with the decomposition of the treatment effect, our results suggest that the effect is more prominent when the price gap between platforms is small and externalities between users or between goods offered by Steam are high. Taken together, our study unveils a unique aspect of strategic promotion in the context of platform competition and highlights the crucial role of externalities in shaping the competitive landscape of digital platforms.

We consider a joint admission and aggregate service rate control problem in a service system. Admission control involves deciding which arriving customers to admit and which to reject. Aggregate service rate control focuses on determining the staffing level and/or service rate for the system. We consider a general reward structure and a convex service cost structure to capture many variations that arise in practice. We show that, under specific structural properties of the revenue and cost functions, the joint optimization of admission and aggregate service rate decisions can be analyzed to produce interesting and implementable policies. Specifically, for systems with a linear service cost structure, we identify a critical operational threshold for the arrival rate. For arrival rates below this threshold, the optimal policy is to close the system and reject all customers. Above this threshold, the optimal policy is to admit all customers and choose an aggregate service rate that depends on the actual arrival rate. In contrast, for systems with a strictly convex service cost structure, under certain conditions, we identify two operational thresholds for the arrival rate. When the arrival rate is below the lower threshold, it is optimal to close the system. It is optimal to admit all customers and choose an arrival-rate-dependent aggregate service rate for arrival rates between the two thresholds. Above the higher threshold, the optimal admission rate and the optimal aggregate service rate do not change any further. We further demonstrate the value of joint optimization by comparing it with two natural benchmarks—optimizing admission rate alone and optimizing aggregate service alone. In stationary arrival settings, we show that joint optimization not only informs the critical decision of whether to operate but also shows that optimizing admission rate or aggregate service rate alone can lead to significant profit losses. We then extend the analysis to nonstationary arrivals with real-world call center data, which further demonstrates the practical value of joint optimization.

Blood type substitution is an effective policy for improving the blood supply chain resilience in responding to shortages at hospitals. However, while compatibility of certain blood types provides an opportunity to better manage volatility in supply (donation) and demand (transfusion), in practice, it remains challenging to decide whether to use a compatible blood unit from the on-hand inventory (substitution) or to place an urgent or emergency order. Available evidence indicates that, in Australia, current practices for deciding on blood type substitution over emergency orders result in a significant imbalance in supply and demand for the most widely compatible units: O-negative blood units. This challenge is particularly critical in blood inventory management and represents a significant research opportunity for the operations management community. It can be addressed using state-of-the-art methodologies in data-driven decision making and perishable inventory under stochastic supply and demand. This study proposes a stochastic optimization model based on sample-average approximation (SAA) to aid the blood type substitution decisions at hospitals. We take into account the historical demand data, the type and age of blood units in the inventory, as well as the shelf-life variability of the units arriving at hospitals. We compare the SAA results with several policies, including a single-period myopic model, no-substitution, over-order practice, and an empirical guideline. Results indicate that the proposed SAA model outperforms all the above policies. It thus creates significant opportunities for improvement in the management of blood units, as it reduces the imbalance in the supply and demand of O-negative blood units, allowing practitioners to benefit from substitution while avoiding the over-ordering of O-negative units.