Knowledge that Transforms

Problem definition: In emerging markets, sharing economy platforms that connect customers with independent service providers often operate in environments with low digital literacy and small, fragmented demand. To address these challenges, such platforms often rely on human intermediaries, known as booking agents, to collect demand from individual customers and submit the aggregated demand on the platform. The presence of such agents requires the platform to set not only the customer price and provider wage but also the agent wage to coordinate supply and demand. This paper analyzes the platform’s pricing and wage decisions and examines how the presence of booking agents affects the surplus of providers, customers, and the platform. Methodology/results: We model a platform involving providers, customers, and booking agents and characterize the platform’s optimal price, wages, and equilibrium outcomes. Our analysis yields several actionable insights. First, a larger provider pool can make it optimal for the platform to raise agent wages while lowering customer prices, even if this combination may reduce the platform’s commission. Second, platforms may find it optimal to pair surge pricing with increased agent wages and decreased provider wages, departing from the conventional strategy of pairing surge pricing with increased provider wages. Finally, whereas the presence of booking agents increases provider earnings by enabling more demand to be served, it may not always benefit customers or the platform. Nevertheless, these agents lead to a “win-win-win” outcome when agents’ demand aggregation cost is moderate or when providers incur high fixed costs in serving colocated customers. Managerial implications: Our findings highlight how platforms should respond to different supply and demand conditions in the presence of booking agents and inform when the use of booking agents generates value for all stakeholders. Our insights have informed changes in the practice of our industry partner, Hello Tractor.

Problem definition: Interest rates on loans are often influenced by market prospects. Under asymmetric information, firms may attempt to signal strong prospects to lenders by over-borrowing. However, publicly revealing confidence in the market can also intensify competition. Motivated by these observations, we examine a firm’s financing and information disclosure strategy. Methodology/results: We develop a game-theoretical model where a firm with private information about the market prospect competes in quantity against a representative competitor. The firm has a limited amount of internal capital and must borrow from lenders to finance production. We show that when borrowing information is publicly accessible, the firm’s strategy depends on its capital needs and the level of competition. If capital needs are high and competition is low, the firm over-finances under strong market prospects to secure a lower interest rate. Conversely, if capital needs are low and competition is high, the firm under-finances under weak market prospects to reduce competitive pressure. Interestingly, in other scenarios, these opposing incentives neutralize each other, leading to a first-best outcome. We further explore private financing, where borrowing information remains undisclosed, forcing the competitor to rely on prior market information. We find that public financing generally dominates private financing under intense competition, but it can also be advantageous when competition is low and capital needs are high. Managerial implications: Our findings suggest that, under information asymmetry and competition, external borrowing can sometimes be beneficial by creating a counterbalancing force, and a slight increase in competition is not always detrimental when financing needs exist. History: This paper was selected as part of the 1RR initiative between the M&SOM Journal and the MSOM Society. This paper was part of the 2024 MSOM Service Operations SIG Conference.

Problem definition: For professional products, such as musical instruments and sports gear, a consumer’s quality preference is positively associated with the consumer’s expertise level. A novice who initially chooses a low-quality product over a high-quality one may have incentive to upgrade if experiencing expertise advancement after professional training. This paper examines a firm’s strategies to sell professional products in two periods, training and posttraining, between which consumers’ expertise levels advance with uncertainty. Methodology/results: Using a game-theoretic model, we find that consumer uncertainty about expertise advancement allows the firm to expand product line and implement expertise-based intertemporal segmentation in selling a line of professional products. In addition, the firm can implement ex ante market segmentation, selling products of different qualities to consumers with different quality preferences before their training starts. Our analysis reveals the innate conflict between these two types of segmentation so that the firm may forego the opportunity of market segmentation and implement intertemporal segmentation only. Interestingly, offering a trade-in credit facilitates intertemporal segmentation and encourages the firm to offer a product line, whereas a buyback program facilitates market segmentation instead. In addition, the total consumer surplus increases whenever trade-in credit enables upgrading, and otherwise, it deteriorates or remains the same. Finally, we find that the firm is more likely to offer a product line when selling to myopic consumers, and consumer myopia may hurt firm profit. Managerial implications: Marketers of professional goods can strategically facilitate and meanwhile, profit from consumer upgrading. Also, whereas a buyback guarantee is commonly used in a regular goods market to cope with buyer uncertainty, a trade-in program is often a better option in a professional goods market.

Problem definition: Motivated by online dating platforms, we study the problem of selecting which subset of profiles to display to each user in each period. Users observe the profiles set by the platform and decide which of them to like, and a match occurs if and only if two users mutually like each other, potentially across different periods. The platform aims to maximize the expected number of matches produced over the entire time horizon, and users’ behavior—captured by their like probabilities—may depend on their history. Methodology/results: We develop a general theoretical model that captures the dynamic, two-sided nature of the problem and the influence of users’ past experiences on their future behavior. We focus on one-lookahead policies and propose the Integral Dating Heuristic (DH-int), providing formal performance guarantees: DH-int achieves a uniform [Formula: see text] approximation across all platform designs under reasonable assumptions. Our empirical analysis, using proprietary data from a major U.S.-based dating app, confirms that DH-int consistently outperforms other benchmarks such as Greedy, Perfect Matching, and Dating Heuristic (DH) and approaches the theoretical upper bound across multiple platform designs and variants of the history effect. The superior performance of DH-int is driven primarily by its careful balancing of initial and follow-up interactions, which accounts for the two-sided nature of the market. Managerial implications: DH-int offers a simple, implementable framework that can substantially improve matching outcomes. Our results provide actionable guidance for curated dating platforms on sequencing, allocation, and leveraging behavioral dynamics. More broadly, the insights extend to other complex, dynamic, two-sided marketplaces—such as freelancing, ride-sharing, and accommodation platforms—where careful sequencing and allocation decisions are critical to optimizing overall outcomes.

Problem definition: Addressing capacity management within a multifaceted network of resources is a critical challenge in service operations management. As resources are often shared to serve multiple classes of customers in such systems, customer routing often depends on the congestion levels of these resources, which in turn are affected by customer routing, creating a feedback loop. Consequently, when evaluating the impact of substantial capacity changes in the network, one must account for the complexity resulting from resource sharing, endogenous routing policies, and the feedback loop between routing policies and congestion levels that has the potential to alter the equilibrium of the system. This complexity renders conventional approaches insufficient for an accurate assessment. Methodology/results: To tackle these challenges, we develop a structural estimation approach that relies on two key components. First, we estimate the routing policy via a choice model that allows the routing policy to depend on not only the focal resource’s utilization but all connected resources’ utilization. We adopt a control function approach with instrumental variables to estimate the loads’ effect in routing without bias. Second, we incorporate the estimated routing policy into a queueing network model that captures the detailed system dynamics and evaluates the equilibrium performance of the entire network. We apply our approach to the specific empirical setting of the hospital inpatient ward network. We show that our proposed approach outperforms two alternative models and highlight the importance of accounting for network equilibrium effects when evaluating substantial capacity changes. Managerial implications: We provide prescriptive capacity allocation recommendations to hospital managers. More generally, our findings underscore the importance of a comprehensive understanding of the interdependencies between customer routing decisions and the levels of congestion present at various resources, shedding light on broader strategies for improving the operational performance of service networks.

Problem definition: We collaborate with an industrial partner whose supply chain uses multiple tiers, locations, and shipping speeds to efficiently serve customers. In practice, our partner also faces the possibility of upstream disruptions, which limit inventory availability. We model these key features of our partner’s network as a multiechelon distribution system (central warehouse and retailers) with expediting and disruptions. Methodology/results: We prove a novel stochastic program lower bound on optimal cost in this model and use this program to develop a heuristic base-stock policy. Our analysis demonstrates that there is a pronounced benefit from centralized inventory (i.e., holding inventory at the central warehouse) in distribution systems with expediting and disruptions as it can be used to both clear backlogs through expediting and hedge against future disruptions. Further, in the disrupted mode, we provide a simple criterion to determine when decentralization (i.e., holding inventory at the retailers) is preferred over complete centralization. Then, we validate our policies using data from our partner’s nationwide distribution network in the United States. Managerial implications: We provide novel inventory policies for managing a distribution system with expediting and disruptions that are understandable and implementable in practice. Our analysis provides the insight that facilitating the right level of central warehouse inventory is a critical hedge for improving performance in these systems. Finally, our industrial partner’s data suggest that our policies can provide significant cost savings in practice.

Problem definition: In the context of outpatient care, physicians often decide at the end of a consultation session whether to schedule follow-up appointments for patients with likely future needs. Those appointments are referred to as prioritized follow-up (PFU) appointments. We study mechanisms that encourage physicians to schedule the optimal quantity of PFUs, aiming to enhance continuity of care (COC), minimizing no-shows and late cancellations. Methodology/results: Utilizing both empirical analysis and modeling, this paper examines strategies for enhancing COC within an appointment scheduling framework. It presents empirical evidence indicating that a greater frequency of PFUs is associated with improved COC. Subsequently, a queueing model is introduced that delineates the impact of PFU appointments on revenue generation and COC levels. The model reveals a discrepancy between physician and health system preferences regarding the number of PFUs with doctors inclined to schedule fewer and health systems favoring more. We apply a principal–agent model to examine the optimal decisions of both the clinic (principal) and the doctor (agent). For situations involving symmetric information, we identify a performance-based payment structure that effectively aligns the incentives of both stakeholders. When information is asymmetric, we theoretically compare four distinct contract types and identify the most promising candidate. Managerial implications: Our findings suggest that health systems ought to implement incentive schemes that reward physicians for a higher proportion of PFUs in their appointment schedules. Such rewards are more cost-effective than those based on the aggregate number of PFU appointments. History: This paper was selected as part of the 1RR initiative between the M&SOM journal and the MSOM Society. This paper was part of the 2024 MSOM Service Operations SIG Conference.

Problem definition: With the advance of data analytics, many disease prediction models have been developed with the intent of detecting diseases earlier and improving patient outcomes through earlier treatment. The operationalization of interventions and care based on these predictive models is critical to attaining these goals. We study the real-world effects of a machine learning-guided colorectal cancer screening outreach program deployed at a health system in Pennsylvania. Methodology/results: Using a regression discontinuity design based on the predicted risk score for having cancer, we find that the program increases the likelihood of colonoscopy uptake in three and six months by 6.0 percentage points (214% increase relative to the control sample within the bandwidth) and 6.9 percentage points (117% increase), respectively. Importantly, we also find significant effects on mortality. We estimate that the program decreases two-year mortality by 6.2 percentage points (43% decrease). Managerial implications: Our finding suggests that a proactive cancer screening outreach program where individuals are selected for intervention based on a machine learning algorithm could significantly improve patient outcomes in addition to achieving higher disease detection rates. Our analysis demonstrates an analytical framework for rigorously evaluating machine learning-aided outreach programs for other cancers and diseases. Establishing unbiased estimates of the impact of machine learning-aided screening outreach is critical for capacity planning of screening resources, such as colonoscopies. History: This paper has been accepted as part of the 2025 Manufacturing & Service Operations Management Practice-Based Research Competition.

Problem definition: Following recent legislation, private insurers participating in Medicare Advantage (MA) are allowed to expand their set of supplementary services to include benefits that are not primarily health related. This policy is meant to lead to lower costs and better care, but questions have been raised about how this affects the incentives of private insurers. We investigate the implications of the policy for MA offerings, beneficiary enrollment, and social welfare. Methodology/results: We develop a game-theoretical model of the interactions between a social planner, an MA insurer, and beneficiaries. We posit that, even after risk-adjustment based on clinical factors, the beneficiary population is heterogeneous in their vulnerability to severe health events due to social determinants of health (SDOH). Owing to this heterogeneity and the broadened remit of the new legislation, the insurer can therefore introduce (i) preventive supplementary services, reducing the risk of severe health events and therefore mostly benefiting vulnerable beneficiaries, or (ii) general well-being supplementary services that entail the same level of non-health-related utility for all beneficiaries. We demonstrate that the policy change incentivizes the insurer to provide preventive supplementary services—which have the largest health impact—if and only if it receives high capitation payments for enrolled beneficiaries. Otherwise, the insurer offers general well-being supplementary services at prices tailored to attract less vulnerable beneficiaries. Moreover, for intermediate values of the capitation payment, the insurer enrolls fewer beneficiaries than under status quo policies—highlighting how the policy change can backfire by exacerbating favorable beneficiary selection (cream skimming) issues inherent in capitation payment systems. Finally, we investigate whether the social planner can mitigate favorable selection by adjusting the capitation payment. We find that the planner may need to choose between inducing improved health outcomes at a high cost (via preventive services) and higher overall welfare (via general well-being services). Managerial implications: While allowing private insurers to offer a wider range of supplementary services has the potential to result in more preventive offerings (thereby lowering treatment costs), policymakers should be aware of the conditions under which the policy might, instead, increase cream skimming and reduce social welfare. Avoiding such outcomes may be costly.

Problem definition: This paper investigates a periodically reviewed distribution inventory system where a central warehouse replenishes multiple retailers facing uncertain demand. Only moment information about demand at each retailer is available, and unmet demand is backlogged. Methodology/results: We develop a robust multiperiod inventory model for the system based on the central limit theorem–based uncertainty set and transform the inventory planning problem into a transportation problem. We characterize two conditions under which a Monge sequence exists for the transportation problem and derive the optimal ordering decisions for the robust inventory model. Building on the robust optimal policy structure, we propose a priority-based inventory policy with look-up-to-k-period reservation. Under this policy, each retailer maintains both an order-up-to level and a reservation target based on the number of periods each retailer looks ahead. Managerial implications: Numerical experiments show that our policy outperforms the other benchmark policies from the literature. The advantage is particularly pronounced under robust performance measures and with real-world demand data that exhibit high variability, skewness, and tail risk. This highlights the strong ability of our policy to handle extreme cases in real-world data sets.