As Charles Darwin’s evolution theory reveals, it is not the strongest species that survive, but those most responsive to change. This principle also applies in the realm of operations management, where managers shall creatively redesign operations to address new challenges. This dissertation presents three cases where renovating traditional operations cost-effectively solves emerging problems, including fraudulent reviews on online platforms (Chapter 1), inefficient strategy design of advertisers (Chapter 2), and inadequate user participation in global procurement initiatives (Chapter 3). I demonstrate that such a practice not only enhances operational efficiency but promotes social welfare. The first two chapters examine operational renovation in the private sector, while the third focuses on the public sector. Specifically, Chapter 1 investigates sellers’ review manipulation on e-commerce platforms and shows that platforms may not be as committed to combating fake reviews as they claim to be. To mitigate this problem, I craft a game-theoretic model and illustrate that restructuring return policies – an essential, long-established operation – can inhibit review manipulation. Chapter 2 analyzes geofencing, an emerging advertising strategy that enables advertisers to send ads to consumers within a virtual fencing zone. While extant literature shows the usefulness of geofencing, the optimal implementation of the strategy remains unclear. Therefore, I analytically examine the optimal operations of geofencing. The findings suggest that the typical practice of setting the geofence around the advertiser’s store is not cost-efficient. Advertisers shall think outside of the box and consider placing the fencing zone elsewhere. My proposed geofencing location and radius could increase resource utilization, advertising efficacy, and consumer welfare. Chapter 3 switches the focus to the public sector, addressing the unaffordability of health products in low- to middle-income countries (LMICs). Social planners have managed procurement pools to help LMICs access health products, yet countries’ willingness to join the pool can vary greatly. A lack of country participation would jeopardize the success of pooled procurement. To encourage more countries to join, I design a procurement mechanism that considers countries’ heterogeneous preferences, disease burdens, and ability to pay. This proposed mechanism, with an appropriately designed subsidy plan, could maximize the aggregate social welfare.

Nonprofits and humanitarian organizations play a critical role in the modern world. Yet, to operate sustainably, they often encounter challenges including financial insecurity and operational obstacles. My dissertation investigates nonprofits' decisions and strategies for delivering sustainable services from the perspectives of financial security and operations in short- and long-term horizons.The first chapter is focused on the role of governance quality in nonprofits' donation income. Donors, generally, support charities that maintain higher program spending ratios (PSR). Yet, PSR does not reflect charities' actual social impact, and a focus on PSR may eventually limit their capacity in providing humanitarian aid. Since 2008, as a result of a policy change by the U.S. Internal Revenue Service, nonprofits are able to better display their governance quality. My empirical investigation shows that governance quality is now an important factor in driving donations to nonprofits, although PSR still remains a key driver. Results suggest that nonprofits should consider improving their governance quality in their strategies for securing donation income, although that may lead to lower PSRs. Pressures resulted from the focus on PSR encourage nonprofits to prioritize strategies that enable them to report higher program expenses. In the second chapter, I empirically examine one of these strategies, grant provision, that allows nonprofits to increase their reported program expenses without having to spend their funds on their own programs. I find that providing grants to other organizations enables nonprofits to earn more revenue and make a bigger social impact in the long term, but this strategy increases the administrative burden needed to make an impact. Given the challenges in coordination and lack of effective coordinated response in humanitarian operations, in the third chapter, I develop a non-cooperative game theoretical model to analyze horizontal coordination among non-governmental organizations in disaster relief operations in centralized and decentralized models. I show that coordination does not always maximize social welfare, and time inefficiencies due to bureaucracies involved in coordination mechanisms are substantial obstacles against higher levels of coordination, especially in urgent response operations. I also show that decentralization of coordination mechanisms increases both coordination levels and social welfare.

In the first chapter, I consider a capacity and price bounded profit maximization problem in which a firm determines prices of multiple substitutable products when the supply or capacity of the products is limited and the prices are bounded. This problem applies broadly to many pricing decision settings such as for hotel rooms, airline seats, fashion, or other seasonal retail products, as well as any product line with shared production capacity. In this paper, I characterize structural properties of the constrained profit maximization problems under the Multinomial Logit (MNL) model and the optimal pricing solutions, and present efficient solution approaches. In the second chapter, I consider a data-driven profit maximization problem in which a firm determines the prices of multiple substitutable products. This problem applies broadly to many pricing decision settings such as for hotel rooms, airline seats, fashion, or other seasonal retail products. A typical data-driven optimization problem takes a two-step approach of parameter estimation and optimization for decisions. However, this often returns a suboptimal solution as the estimation error due to the variability in data impacts the quality of the optimal solution. I present the relationship between estimation error and quality of the optimal solution and provide a possible way to reduce the impact of the error on the optimal pricing decision under the MNL model. In the last chapter, I consider a facility layout design problem of a semiconductor fabrication facility (FAB). In designing a facility layout, the traditional approach has been to minimize the flow-weighted distance of materials through the automated material handling system (AMHS). However, distance focused approach sometimes yields one major issue, traffic congestion, that there is a question if it is truly a good criterion to design a layout. In this study, I try to understand what makes such congestion by analyzing the system dynamics and propose another approach with a concept of ``balancing the flow" that focuses more on resolving the congestion. Finally, I compare the performance of the two methods through the simulation of semiconductor FAB layouts.

Platform business models have become pervasive in many aspects of the economy,particularly in the areas experiencing rapid growth such as retailing (e.g., Amazon and eBay) and last-mile transportation (e.g., Instacart and Amazon Flex). The popularity of platform business models is, in part, due to the asset-light prospect which allows businesses to maintain flexibility while scaling up their operations. Yet, this ease of growth may not necessarily be conducive to viable outcomes. Because scalability in a platform depends on the intermediary’s role it plays in facilitating matching between users on each side of the platform, the efficiency of matching could be eroded as growth increases search frictions and matching costs. This phenomenon is demonstrated in recent studies on platform growth (e.g. Fradkin, 2017; Lian and Van Ryzin, 2021; Li and Netessine, 2020). To sustain scalability during growth, platforms must rely on effective platformdesign to mitigate challenges arising in facilitating efficient matching. Market design differs in its focus between retail and last-mile transportation platforms. In retail platforms, platform design’s emphasis is on helping consumers navigate through a variety of product offerings to match their needs while connecting vendors to a large consumer base (Dinerstein et al., 2018; Bimpikis et al., 2020). Because these platforms exist to manage two-sided demand, scalability depends on the realization of indirect network economies where benefits for users to participate on the platforms are commensurate with the size of users on the other side (Parker and Van Alstyne, 2005; Armstrong, 2006; Rysman, 2009). Thus, platform design plays a critical role in the realization of indirect network economies on retail platforms. Last-mile transportation platforms manage independent drivers on one side andretailers on the other, both parties holding flexibility in switching between platforms. High demand for independent drivers along with their flexibility in work participation induces platforms to use subsidies to incentivize retention. This leads to short-term improvements in retention at the expense of significant increases in platforms’ compensation costs. Acute challenges to driver retention call for effective compensation strategies to better coordinate labor participation from these drivers (Nikzad, 2017; Liu et al., 2019; Guda and Subramanian, 2019). In addition to driver turnover, retailers’ withdrawal can undermine the operating efficiency of last-mile transportation platforms (Borsenberger et al., 2018). This dissertation studies platforms’ scalability and operational challenges faced by platforms in the growth.

The global population is expected to reach 10.5 billion by 2050. With the increase in population, food production needs to increase by at least 70% in 2050. This would require a several-fold increase in food production. However, scarcity in land availability, a falling water table, weather variability, and an increase in the cost of agricultural operations have made this difficult. The gap between food supply and demand could be minimized if food losses are reduced during production, post-harvest activities, and food waste during consumption. This dissertation focuses on food loss (FL) by growers and food-waste (FW) by households. Specifically, the dissertation first, investigates the impact of vertical coordination on FL in India. Secondly, the dissertation examines the impact of offline and online shopping on FW by American households. The FL study uses farm-level data from India and a novel estimation method in the literature. Findings show that agribusiness firms rejected a significant quantity of the product due to quality standards. The amount of produce rejected was directly impacted by labor and transportation costs. Modeling and simulating the effects of labor and transport costs show that lowering labor and transport costs for the smallholder growers would reduce FL. The FW study uses scanner data of a popular retailing chain in the United States. Using the behavior of over-purchasing of impulse products and machine learning approach, the predict the over-purchasing of impulse products across online and offline (grocery stores) channels. The study finds that households over-purchase 29% more of impulse products (danish pastries, sweet bread, and cakes) when shopping online compared to offline shopping. The dissertation provides two critical insights related to the decision-making process of growers and grocery shoppers. First, growers' decision on reducing FL is related to the quantity of produce rejected by contracting firms and selling produce in the spot markets. Second, FW is significantly related to a grocery shopper’s choice of a shopping channel and the decision on how much to purchase.

The shift in focus of manufacturing systems to high-mix and low-volume production poses a challenge to both efficient scheduling of manufacturing operations and effective assessment of production capacity. This thesis considers the problem of scheduling a set of jobs that require machine and worker resources to complete their manufacturing operations. Although planners in manufacturing contexts typically focus solely on machines, schedules that only consider machining requirements may be problematic during implementation because machines need skilled workers and cannot run unsupervised. The model used in this research will be beneficial to these environments as planners would be able to determine more realistic assignments and operation sequences to minimize the total time required to complete all jobs. This thesis presents a mathematical formulation for concurrent scheduling of machines and workers that can optimally schedule a set of jobs while accounting for changeover times between operations. The mathematical formulation is based on disjunctive constraints that capture the conflict between operations when trying to schedule them to be performed by the same machine or worker. An additional formulation extends the previous one to consider how cross-training may impact the production capacity and, for a given budget, provide training recommendations for specific workers and operations to reduce the makespan. If training a worker is advantageous to increase production capacity, the model recommends the best time window to complete it such that overlaps with work assignments are avoided. It is assumed that workers can perform tasks involving the recently acquired skills as soon as training is complete. As an alternative to the mixed-integer programming formulations, this thesis provides a math-heuristic approach that fixes the order of some operations based on Largest Processing Time (LPT) and Shortest Processing Time (SPT) procedures, while allowing the exact formulation to find the optimal schedule for the remaining operations. Computational experiments include the use of the solution for the no-training problem as a starting feasible solution to the training problem. Although the models provided are general, the manufacturing of Printed Circuit Boards are used as a case study.