No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
On-Demand Service Platforms
Abstract
An on-demand service platform connects waiting-time-sensitive customers with independent service providers (agents). This paper examines how two defining features of an on-demand service platform—delay sensitivity and agent independence—impact the platform’s optimal per-service price and wage. Delay sensitivity reduces expected utility for customers and agents, which suggests that the platform should respond by decreasing the price (to encourage participation of customers) and increasing the wage (to encourage participation of agents). These intuitive price and wage prescriptions are valid in a benchmark setting without uncertainty in the customers’ valuation or the agents’ opportunity costs. However, uncertainty in either dimension can reverse the prescriptions: Delay sensitivity increases the optimal price when customer valuation uncertainty is moderate. Delay sensitivity decreases the optimal wage when agent opportunity cost uncertainty is high and expected opportunity cost is moderate. Under agent opportunity cost uncertainty, agent independence decreases the price. Under customer valuation uncertainty, agent independence increases the price if and only if valuation uncertainty is sufficiently high.
The online appendix is available at https://doi.org/10.1287/msom.2017.0678 .
... On-demand platforms rely on algorithms to coordinate human resources (e.g., drivers) and to provide timely services (Möhlmann et al., 2021). These advanced algorithms have revolutionized on-demand services by effectively matching consumer demands with available services (Feng et al., 2021;Taylor, 2018). However, maintaining delivery reliability remains challenging (Dai & Liu, 2020). ...
... Further, although previous studies have assessed related issues, such as delay sensitivity and waiting time, in on-demand services, few empirical studies have examined the factors influencing delivery reliability (i.e., low delivery delay) (Bai et al., 2019;Shang & Liu, 2011;Taylor, 2018;Yan et al., 2021). Hence, the literature has called for studies focused on "identifying and quantifying key factors that affect delivery performance" (Mao et al., 2022). ...
... Among on-demand services, maintaining delivery reliability is challenging (Cho et al., 2008;Lee et al., 2015), as consumers using these platforms are highly sensitive to delays and waiting time (Taylor, 2018;Xu et al., 2021;Yan et al., 2021). Therefore, on-demand services have stringent time requirements (Allen et al., 2018;Bai et al., 2019;Taylor, 2018), especially for on-demand meal deliveries, due to the perishable nature of the products (Allen et al., 2018) and the high impatience of hungry customers (Mao et al., 2022). ...
A surge in technological advancements and innovations has spurred the rise of on‐demand meal delivery platforms. Despite their widespread appeal, these platforms face two critical challenges (i.e., order batching and demand allocation) in effectively managing the delivery process while maintaining reliability. In response, this study aims to address these two challenges by examining the effects of real‐time demands and restaurant density on delivery reliability, as well as how the type of driver (i.e., in‐house versus crowdsourced drivers) moderates these effects. We evaluated our model with a unique dataset obtained from one of the top three on‐demand meal delivery platforms in China, and our research sheds light on several key findings. Specifically, our study finds inverted U‐shaped relationships between real‐time demands and delivery reliability and a positive relationship between restaurant density and delivery reliability. In addition, it reveals that crowdsourced drivers perform better than in‐house drivers under high real‐time demands. This study contributes to the literature by clarifying how delivery reliability can be influenced by real‐time demands and restaurant density. The results offer important implications for on‐demand meal delivery platforms to improve delivery performance and allocate demands amid complicated market conditions.
... This study is most closely related to three streams of the service operations management literature that employs a combination of the game-theoretic framework and the queueing theory to analyze the operational implications of pricing strategies, online healthcare service, and channel competition (Betcheva et al. 2021;Keskinocak and Savva 2019;Dai & Tayur, 2021;Fainman & Kucukyazici, 2020).Query In the first research stream, numerous works have been examined in terms of service pay rate and price rate in the on-demand service 2 and two-tier healthcare markets. For example, based on an on-demand service platform, Taylor (2018) and Bai et al. (2019) characterize the optimal price that the platform charges its customers and the optimal wage the platform pays its independent providers to coordinate endogenous demand with endogenous supply; Benjaafar et al. (2022) examine the potential effects of price and wage on labor welfare with different sizes of the labor pool. In a twotier healthcare system that is composed of public hospitals and private hospitals, the pricing policies of private hospitals have been extensively studied in the literature. ...
... where β is a constant that is increasing in the number of service providers and satisfies β ≥ 1 (Benjaafar et al., 2022;Taylor, 2018). In a single-server case, when the patient arrival rate follows a Poisson process and the service times are exponentially distributed, the wait time is exactly equal to 1/S M k + 1/(S M k − D M k ). ...
... Taylor (2018) indicates that the on-demand healthcare service is designed to particularly serve those patients with extremely high waiting cost, and who require immediate healthcare services. ...
Motivated by the recent emergence and rapid growth of online healthcare platforms, this paper investigates its impacts on the healthcare ecosystem by taking into consideration of heterogeneous patients, homogeneous service providers and profit-driven hospitals. A game-theoretical model with a queueing framework is developed, and two healthcare systems are considered, including a monopolistic system with either an offline, traditional hospital, or an online, Internet hospital, and a duopolistic system with both kinds of hospitals. Our study finds that in both systems a large population of patients always increases the healthcare supply and demand and the hospital utilization, which leads to an improvement not only in the hospital profitability, but also in the welfare of both patients and service providers. When considering a monopolistic system only, our study suggests that the monopolistic Internet hospital outperforms its counterpart in hospital utilization only when the wait cost is not high (i.e., patients are less sensitive to wait time), especially when the wait cost is medium and the patient-provider ratio is small. When considering a duopolistic system, this study concludes that the decisions of patients, service providers, and hospital managers of both hospitals are interrelated, and that a large population of service providers increases its hospital’s supply, demand, and wage rate, while decreases its wait times. When comparing the two monopolistic with the duopolistic system, the latter outperforms and improves every entity’s utility (i.e., patients, hospital, and society) when there is abundance of service providers with more being allocated to the Internet hospital, whereas the monopolistic system dominates when the patient population is large, and more providers are allocated to the traditional hospital.
... Simultaneously, it offers individual compensation to offset the inconveniences, incurred by fulfilling the delivery task. This assumption is in line with the literature, where higher compensations typically increase the willingness of ODs to accept these offers (see, e.g., Cachon et al. 2017;Taylor 2018;Yildiz and Savelsbergh 2019). ...
... This relationship is explored in an aggregated context, often without investigating the specifics of individual OD behaviors. Key contributions in this area include works by Cachon et al. (2017), Kung and Zhong (2017), Taylor (2018), Qi et al. (2018), Yildiz and Savelsbergh (2019) and Cao et al. (2020). Notably, Cao et al. (2020) speculate on the potential for significant savings through an online compensation framework, particularly when combined with their dynamic task assignment framework. ...
Amid the rapid growth of online retail, last-mile delivery faces significant challenges, including the cost-effective delivery of goods to all delivery locations. Our work contributes to this stream by applying dynamic pricing techniques to effectively model the possible involvement of the crowd in fulfilling delivery tasks. The use of occasional drivers (ODs) as a viable, cost-effective alternative to traditional dedicated drivers (DDs) prompts the necessity to focus on the inherent challenge posed by the uncertainty of ODs’ arrival times and willingness to perform deliveries. We introduce a dynamic programming framework that offers individualized bundles of a delivery task and compensation to ODs as they arrive. This model, akin to a reversed form of dynamic pricing, accounts for ODs’ decision-making by treating their acceptance thresholds as a random variable. Therefore, our model addresses the dynamic and stochastic nature of OD availability and decision-making. We analytically solve the stage-wise optimization problem, outline inherent challenges such as the curses of dimensionality, and present structural properties. Tailored to meet these challenges, our approximation methods aim to accurately determine avoided costs, which are a key factor in calculating optimal compensation. Our simulation study reveals that the savings generated by involving ODs in deliveries can be significantly increased through our individualized dynamic compensation policy. This approach not only excels in generating savings for the firm but also provides a utility surplus for ODs. Additionally, we demonstrate the applicability of our approach to scenarios with time windows and illustrate the trade-off that arises from time window partitioning.
... In addition to time sensitivity, Bahrami et al. (2023) studied the problem under the assumption that customers, along with food suppliers and couriers, also show sensitivity to pricing policies. For those interested in exploring the topic further, there is also a body of research dedicated to pricing strategies in on-demand food delivery services (Taylor, 2018;Du et al., 2023a,b;Behrendt et al., 2024;Ji et al., 2024;Zhou et al., 2024). ...
In this paper, we address the complex, dynamic, multi-stakeholder problem of on-demand food delivery, where customers at different locations continuously place orders to various restaurants for courier delivery. To tackle this challenge, we formulate a novel Stochastic Dynamic Order Dispatching Problem (SDODP) as a Markov Decision Process (MDP), integrating dynamic information from pending orders, courier statuses, and aggregated order features into the state representation. We propose a multi-step deep reinforcement learning (MDRL) approach based on Proximal Policy Optimization (PPO) with a discrete action space to manage order assignments in the first step. In the next step, a novel Matching Degree function evaluates the compatibility between orders and couriers by balancing order importance, time remaining until the deadline, and courier efficiency. Finally, we employ a dynamic heuristic routing procedure to optimize courier delivery routes. We perform extensive data analysis to generate realistic inputs, including spatio-temporal order patterns, geographical clustering of customers and restaurants, courier speed distributions, and other key parameters. To evaluate the effectiveness of our approach, we compare our dynamic MDRL-based policies against multiple static benchmark policies, each developed by incorporating a fixed fraction of assigned pending orders and distinct order selection criteria (e.g., First-Come-First-Served). Computational results demonstrate the superiority of our MDRL policies on unseen test cases in terms of total deliveries, pending orders, and overdue orders. Finally, we provide managerial insights based on extensive experiments and a thorough sensitivity analysis of courier numbers and capacities.
... With a globally rising availability of smart devices (Ostrom et al. 2015), a broad spectrum of citizens is continuously accessing cyber-physical ecosystems on a daily basis (Wortmann and Flüchter 2015;Zhang et al. 2012). In these digital ecosystems appear several platform types predominantly building on service exchange (Bartelheimer et al. 2022;Taylor 2018). For instance, not only the message delivery process through social media platforms is a frequently consumed service, but also booking flexible ridesharing services by means of commercial platforms has become omnipresent (Cachon et al. 2017). ...
... By using on-demand technology, Uber fulfils customer requests in real time using independent contractors instead of employees. Platforms differ from the traditional firm-employee model because they do not determine when an individual works or pay a salary (Taylor, 2018). This business model enables the platform to save significantly on payroll costs; they also operate a low asset base (drivers use their own vehicles to provide the service), outsource payment and logistics services (e.g. using Braintree to process payments and Google maps for navigation) and avail of tax havens (Srnicek, 2017). ...
This study investigates the dynamics of worker resistance within the gig economy. Drawing on a combination of 36 qualitative interviews with Uber drivers and a netnographic analysis of the forum Uberpeople.net, we investigate how drivers use digital communities to challenge precarity. Despite Uber's efforts to individualize and control their labour, we reveal that resisting drivers use online forums to share experiences, develop resistance strategies, and foster collective identity. Specifically, we identify three primary mechanisms through which online forums facilitate resistance: (1) fostering in-group solidarity through shared grievances and collective identity formation; (2) enabling information exchange that empowers drivers to navigate and challenge platform constraints; and (3) providing discursive justifications for non-compliance with platform rules. This study contributes to research on labour resistance and algorithmic management by demonstrating how gig workers leverage digital spaces to contest control, highlighting the central role of storytelling and online communities in shaping contemporary labour struggles within the gig economy.
... The study of the challenging economics and operational complexity often associated with LMO-involving a wide variety of actors and combinations of activities that are insourced with those that are outsourced-would profit from theory on business models (e.g., platforms (Taylor 2018) and data-driven models (Rooderkerk et al. 2022)), as well as supply chain integration (e.g., Frohlich and Westbrook 2001) and supply chain complexity (e.g., Bozarth et al. 2009). ...
Last mile operations (LMO), the processes involved in the critical last stage of delivering goods and services, have widespread relevance across major sectors of the economy, including retail, food services, healthcare, humanitarian services, energy distribution, telecommunications, public services, and others. These operations account for a significant portion of the costs, jobs, and economic output in these sectors. Global economic output involving last mile deliveries alone, for instance, is valued at $165 billion per year and is growing at about 10% per year (InsightAce Analytic 2024).
Recent decades have witnessed an acceleration in the rate of evolution of LMO (Agatz et al. 2024; Boutilier and Chan 2022; Boyer and Hult 2005; Dreischerf and Buijs 2022; He and Goh 2022; Lyu and Teo 2022). Technology-driven innovations have catalyzed profound changes in the planning, design, and execution of LMO, with significant implications for the economics of these operations. Extending the last mile to the final user has increased convenience, accessibility, and reliability. Zipline, for example, has introduced drones to safely deliver lifesaving products in remote communities (Ackerman and Koziol 2019). An increasing number of pharmacies in Europe and Africa have been equipped with smart lockers to allow 24/7 access to critical medicines (Gobir et al. 2024). Some innovations leveraging platforms based on smartphone apps have given small corner stores in neighborhoods in cities across Latin America the means to sell and deliver daily groceries and other household staples to local residents (Escamilla et al. 2021). Other innovations, leveraging artificial intelligence, have found applications in vehicle routing tools and warehouse and fulfillment automation (such as Ocado's system (Mason 2019)), track-and-trace systems that provide real-time communications and visibility into delivery processes (such as Instacart and Uber Eats), anticipatory shipping algorithms to move inventories to specific areas ahead of realized demand (Chen and Graves 2021), and integration tools with third-party services (successfully deployed by ClickPost and ShipEngine).
However, considerable challenges remain. For example, because of short time frames and high delivery volumes to many dispersed locations, LMO have little room for human error. Yet, since many firms tend to tap into low-skilled, temporary, or crowdsourced labor to provide these services, there is high variability in performance and worker availability. LMO are also expensive, due in part to rising labor costs, delivery failures, more demanding customers, and vehicle and parking restrictions. Although academic research in LMO has a long tradition in Operations Research (see e.g., Agatz et al. (2011), Otto et al. (2018), Boysen et al. (2019) and Reed et al. (2022)), LMO have barely been considered as an operations problem that requires process understanding and management within a sociotechnical system. The need for this is apparent, as increasing evidence points to managerial, economic, and sociotechnical challenges as major determinants of LMO success. Delivery workers have been noted to largely ignore the recommendations by routing algorithms in urban settings (Liu et al. (2023)); working conditions are an increasing societal and corporate concern; and customer experiences are less than satisfactory in many cases. Further, LMO are associated with negative externalities such as emissions, traffic congestion, and the abuse of public parking space. Operational costs are also very high—often up to a point where LMO are loss-making, such as in grocery home delivery. And, while there have been extensive technological innovations, many seem to fail in scaling at large, which could potentially be due to a poor understanding of the LMO from a process perspective.
We need new research to better understand these challenges, as well as to propose new operational practices and business models based on the application of recent innovations. Such research requires a broadening of the phenomenological and theoretical scope of LMO research beyond traditional work in Operations Research. Theories on innovation applied to Operations Management can offer a valuable foundation to study research questions surrounding the scalability of technologies to support new business models in the last mile (Arthur 1994). Similarly, theoretical models examining technology, productivity, and employment can provide a foundation to understand how innovations can change the nature of work in last-mile settings (Autor et al. 2003; Autor 2015). Additional opportunities also exist to use transaction and information cost theories to understand how technological innovations may change organizational boundaries and the nature of organizations in the last mile (Afuah 2003).
This confluence of innovations in the field, the multidimensional phenomena that determine performance, and the perspectives from theories from the operations management field provide an opportunity to shape a research program in LMO that will benefit from the Operations Management academic community. This was one of the main goals of our call for papers for the special issue on “Innovations, Technologies, and the Economics of Last-Mile Operations.”
Another objective of this special issue was to formalize a research agenda and offer future directions for research to advance our understanding of LMO. To that end, in Section 2, we delve deeper into these operations, their functionalities, distinctive features, and challenges in the context of Operations Management. Then, in Section 3, we expand on research opportunities to tackle the most pressing challenges in LMO and identify knowledge gaps in Operations Management to be addressed in this endeavor. We close in Section 4 with conclusions, recommendations, and potential initiatives to build on the momentum created so far and further advance LMO as a knowledge area within Operations Management. In doing so, we introduce the several papers in the special issue as exemplars of research that can be done in the LMO domain.
... ,Mittal et al. (2021),Taylor (2018),Ghaderi et al. (2022),Cramer and Fikar (2024) 29Castillo et al. (2018),Dayarian and Savelsbergh (2020) 30 Taylor(2018), Qi et al. (2018) 31 Archetti et al. (2016), Arslan et al. (2019), Ghaderi et al. (2022), Behrend et al. (2019), Chen and Chankov (2017) 32 Buldeo Rai et al. (2018), de Oliveira Leite Nascimento et al. (2023), Qi et al. (2018) 33 Fessler et al. (2022), de Oliveira Leite Nascimento et al. (2023) 34 Chen and Chankov (2017), Azcuy et al. (2021), Taylor (2018) 35 Chen and Chankov (2017), Xiao et al. (2023), de Oliveira Leite Nascimento et al. (2023) 36 Taylor (2018) 37 Chen and Chankov (2017), Guo et al. (2019), Voigt and Kuhn (2022), Zhang et al. (2019), Behrend et al. (2019) 38 Dayarian and Savelsbergh (2020), Ji et al. (2020), Li et al. (2019) 39 Carbone et al. (2017), Li et al. (2019) 40 Buldeo Rai et al. (2018), Alharbi et al. (2022), Ermagun and Stathopoulos (2018) 41 Macrina et al. (2020), Ballare and Lin (2020), Ghaderi et al. (2022), Ermagun and Stathopoulos (2018) 42 Ghaderi et al. (2022) 43 Lan et al. (2022), Wicaksono et al. (2022) 44 Rougès and Montreuil (2014), Wicaksono et al. (2022), Le et al. (2019) 45 Le et al. (2019) 46 Li et al. (2019), Rechavi and Toch (2022) 47 de Oliveira Leite Nascimento et al. (2023) 48 Carbone et al. (2017), Azcuy et al. (2021) 49 Devari et al. (2017), Dayarian and Savelsbergh (2020), Castillo et al. (2022a), Bin et al. (2020) 50 Rougès and Montreuil (2014), Azcuy et al. (2021), Devari et al. (2017) 51 Castillo et al. (2018), Chen and Chankov (2017), Kızıl and Yıldız (2023), Castillo et al. (2022b), Arslan et al. (2019) 52 Devari et al. (2017), de Oliveira Leite Nascimento et al. (2023), Kızıl and Yıldız (2023) 53 Chen and Chankov (2017), Castillo et al. (2022a), Ballare and Lin (2020) 54 Castillo et al. (2018, 2022a), Qi et al. (2018) 55 McKinnon (2016), Fessler et al. (2022) 56 Fessler et al. (2022), de Oliveira Leite Nascimento et al. (2023), Azcuy et al. (2021) 57 Qi et al. (2018) 58 Ghaderi et al. (2022), Voigt and Kuhn (2022) 59 Azcuy et al. (2021), Voigt and Kuhn (2022), Ghaderi et al. Buildup of the courier pool and the impact of competition. ...
... Platforms can coordinate supply and demand through two main mechanisms: pricing and assignment. Studies by Bai et al. (2019), Cachon et al. (2017), Taylor (2018), Chen and Hu (2020), and Liu et al. (2022) explore various pricing mechanisms. Alternatively, assignment mechanisms can enhance platform efficiency under constraints like limited capacity (Zhao et al. 2024) or through crowd-sourced delivery with excess capacity (Arslan et al. 2019). ...
The rapid growth of on‐demand meal delivery platforms has heightened competition, making customer retention a critical priority. While prior research on order dispatch algorithms has largely focused on minimizing delivery time or delay, the direct impact of delivery performance on repeat purchases remains underexplored. Using transactional data from an online meal delivery platform in China, we empirically investigate the asymmetric effects of early and late deliveries on customer repurchasing behavior. To address potential endogeneity, we introduce driver experience and local knowledge, two previously overlooked factors in platform algorithms, as novel instrumental variables. The survival analysis shows that late deliveries significantly reduce future orders, while early deliveries provide only limited benefits. Guided by these empirical insights, we develop a simulation‐based evaluation of different order dispatch algorithms, revealing that maximizing future orders, rather than minimizing delivery time or delays, yields the highest future orders. These insights offer actionable recommendations for platform managers, stressing the importance of strategic adjustments in dispatch algorithms and integrating heterogeneous treatment effects into algorithmic design. By merging operational delivery performance with consumer behavior insights through causal inference and optimization, this study provides a novel end‐to‐end framework for creating data‐driven dispatch algorithms that enhance both service efficiency and customer retention.
... A growing body of literature focuses on monetary incentives to increase service providers' Article submitted to Management Science;manuscript no. MS-INS-2025-00086 participation in on-demand platforms (Gurvich et al. 2019, Taylor 2018. Gurvich et al. (2019) show that a firm facing time-varying demand must offer sufficiently high compensation to attract enough gig workers to provide an adequate service level during high demand periods. ...
... On-demand service platforms connect independent workers with consumers seeking time-sensitive services (Bai et al. 2018, Taylor 2018, Benjaafar and Hu 2020, such as ride-hailing and food delivery. Prior studies, particularly within the ride-hailing context, have extensively explored operational challenges faced by these platforms, including issues related to pricing (Cachon et al. 2022, Guda andSubramanian 2019), matching (Bai et al. 2018, Liu et al. 2023) and platform competition (Bernstein et al. 2021, Zhang et al. 2022a. ...
The increasing frequency of traffic accidents involving food delivery drivers has raised public concern about the safety of fast delivery services. In response, food delivery platforms have introduced two corporate social responsibility (CSR) strategies to enhance drivers’ road safety: One in which the platform assumes full responsibility for extending delivery times, and another that empowers consumers to decide on extended delivery times. Additionally, policymakers are mandating that platforms purchase employment injury insurance for their drivers. Despite these efforts, debates continue over how platforms should adjust delivery times to mitigate driver risk and whether new driver protection policies are truly effective. In this study, we employ the Hotelling model to examine a duopoly food delivery market in which consumers are socially responsible, time-sensitive, and exhibit heterogeneous loyalty to competing platforms. Building on the consumer empowerment framework, we investigate the optimal CSR strategy choices for these platforms and assess the resulting impacts on various stakeholders. Our analysis provides three key insights. First, we demonstrate that offering differentiated extended delivery time (EDT) options can mitigate the negative impact of consumer aversion to empowerment on platform profitability, especially in markets with a significant consumer heterogeneity in time sensitivity. Second, we identify a potential all-win CSR outcome, wherein positively shaping consumer attitudes benefits all stakeholders, a possibility often overlooked in practice. Third, we surprisingly find that the widely adopted protective measure, namely the mandatory insurance policy (MIP), does not effectively mitigate driver risk. In contrast, our proposed risk-based MIP (RMIP) addresses this limitation and allows consumers who actively participate in CSR initiatives to benefit from a lower markup on delivery fees. These insights offer a refined analytical framework for understanding consumer empowerment, providing valuable guidance for researchers, platform operations managers and policymakers.
... Kung and Zhong [32] investigate the pricing decisions of a two-side platform with the delivery service considerations. Taylor [33] studies the optimal supply-demand matching in ondemand service platforms with independent service agents and considers the optimal service pricing decision. Most recently, Barenji et al. [34] explore smart platform pricing for e-commerce-based logistics services. ...
This paper investigates using blockchain technology to fight deceptive counterfeits in an electronic commerce environment. Thereby, a two-period pricing model is built under two competitive platforms: a blockchain-based platform which ensures product authentication and provides a higher value to customers but increases customers’ privacy concerns, and its rival (i.e., the traditional platform) in the absence of blockchain implementation which is perceived as having a lower value due to the existence of deceptive counterfeits and thus faces more government enforcement. Customers on both platforms are influenced by the electronic word-of-mouth (eWOM) effect, and customers value a platform more if the platform has more online sales. The two platforms either adopt the fixed pricing scheme or the modifiable pricing scheme and so four possible cases may occur. By deriving the equilibrium of each possible case, we analytically find that the attenuation of consumer privacy concerns, increases in government enforcement efforts, and eWOM can benefit the platform’s adoption of blockchain technology to combat counterfeits, and a strong eWOM effect is conducive to consumers but deteriorates price competition and thus harms both platforms. Whether the pricing schemes enhance the competitiveness of the blockchain-based platform over its rivals depends on the eWOM effect and the advantage gained from adopting blockchain technology.
... They verified that the revenue improvement brought by the dynamic pricing strategy is predominantly driven by an increase in the number of consumers serviced rather than an increase in average price. In addition, researchers investigated the pricing policies considering the riders' and drivers' behavior, e.g., riders' delay sensitivity and drivers' idle-time sensitivity [32,36,37]. ...
On-demand ride-sharing services change our travel behavior, which threatens the survival of taxi services. Motivated by this, this paper examines the impact of on-demand ride-sharing services on taxi services and how to achieve the coexistence of two services from a service quality perspective. This paper analyzes the coexistence condition of two services considering the network effect. First, the profit target for taxi drivers is nonnegative. A Stackelberg model is built in which the taxi service is the leader and the on-demand ride-sharing service is the follower. Then, the reference dependency theory is introduced to modify the profit target of taxi drivers. And the new coexistence condition is compared with the benchmark status. The results find that the coexistence of the two services depends on the type of riders and the quality difference in both cases. When two services and riders are highly heterogenous, two services are more likely to coexist. Services with different qualities could better satisfy the diverse preferences of riders. Considering taxi profit without competition as the profit reference point, the requirement of service quality difference and the diversity of rider composition is increased. In terms of the network effect, the negative network effect is more beneficial to the coexistence of two services.
... Gurvich et al. (2019) studied such a platform and consider self-scheduling agents that decide to work based on expected compensation and their availabilities. Similar works focus on surge pricing to balance demand and supply (Cachon et al., 2017), on the influence of agents' independence and customers' delay sensitivity (Taylor, 2018), and platform commission schemes (Zhou et al., 2019). Similarly to these works, we consider self-scheduling agents as part of our workforce. ...
Nowadays, logistics service providers (LSPs) increasingly consider using a crowdsourced workforce on the last mile to fulfill customers’ expectations regarding same-day or on-demand delivery at reduced costs. The crowdsourced workforce’s availability is, however, uncertain. Therefore, LSPs often hire additional fixed employees to perform deliveries when the availability of crowdsourced drivers is low. In this context, the reliability versus flexibility trade-off which LSPs face over a longer period, for example, a year, remains unstudied. Against this background, we jointly study a workforce planning problem that considers salaried drivers (SDs) and the temporal development of the crowdsourced driver (CD) fleet over a long-term time horizon. We consider two types of CDs, dedicated gig-drivers (DDs) and opportunistic gig-drivers (ODs). While DDs are not sensitive to the request’s destination and typically exhibit high availability, ODs only serve requests whose origin and destination coincide with their own private route’s origin and destination. Moreover, to account for time horizon-specific dynamics, we consider stochastic turnover for both SDs and CDs as well as stochastic CD fleet growth. We formulate the resulting workforce planning problem as a Markov decision process whose reward function reflects total costs, that is, wages and operational costs arising from serving demand with SDs and CDs, and solve it via approximate dynamic programming. Applying our approach to an environment based on real-world demand data from GrubHub, we find that in fleets consisting of SDs and CDs, approximate dynamic programming (ADP)-based hiring policies can outperform myopic hiring policies by up to [Formula: see text] and lookahead policies with perfect knowledge of future information by up to [Formula: see text] in total costs. In the studied setting, we observed that DDs reduce the LSP’s total costs more than ODs. When we account for CDs’ increased resignation probability when not being matched with enough requests, the amount of required SDs increases.
This paper investigates spatial pricing and fleet management strategies for an integrated platform that provides both ride-sourcing and intracity parcel delivery services, leveraging the idle time of ride-sourcing drivers across a transportation network. Specifically, the integrated platform simultaneously offers on-demand ride-sourcing services for passengers and multiple modes of parcel delivery services for customers, including (1) on-demand delivery, where drivers immediately pick up and deliver parcels upon receiving a delivery request; and (2) flexible delivery, where drivers pick up (or drop off) parcels only when they are idle and waiting for the next ride-sourcing order. A semi-Markov process (SMP) model is proposed to characterize the status change of drivers under joint movement of passengers and parcels over the transportation network with limited vehicle capacity, where the service quality of ride-sourcing services, on-demand delivery services, and flexible delivery services are quantified. Building on the SMP model, incentives for ride-sourcing passengers, delivery customers, drivers, and the platform are captured through an economic equilibrium model. Subsequently, we derive the platform’s optimal spatial pricing by solving a nonconvex profit-maximization problem. We establish the well-posedness of the model and introduce a customized algorithm that enhances computation time and numerical stability when determining the platform’s optimal strategies, outperforming the benchmark method. We also validate the proposed model and algorithm through a comprehensive case study of San Francisco. Numerical results indicate that ride-sourcing and parcel delivery services exert both complementary and competitive effects on each other, with the integrated business model’s overall impact hinging on the complex interaction between ride-sourcing orders, on-demand parcel delivery orders, and flexible parcel delivery orders.
Funding: This work was supported by the Research Grants Council of Hong Kong [Grants 16202922 and 26200420].
Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2024.0601 .
Due to driver autonomy, delivery and rideshare platforms in the sharing economy (SE) face uncertainty in service capacity planning. This study investigates how the alignment between the technical and social subsystems within the overall socio‐technical system of SE platforms influences driver engagement. Drawing on socio‐technical systems theory, we conceptualize SE platforms as systems defined by the organizational design constructs of centralization, formalization, and complexity. Using a discrete choice experiment with 100 experienced SE drivers, we examine how platform design attributes affect drivers' utility and engagement probability. Our findings reveal that drivers value flexibility, preferring platforms that offer direct task allocation and single‐task acceptance features. Contrary to expectations, drivers generally favor base pay–oriented remuneration over promotion‐oriented structures. Individual driver characteristics such as SE income dependency and multi‐apping tendencies significantly influence platform preferences. The findings contribute to logistics research by empirically establishing the relationship between platform structure and driver utility and by extending the application of organizational design theory to digital labor platforms. Our results provide insights into how firms can reduce capacity uncertainty by engaging SE drivers within their networks, highlighting the importance of aligning platform design with the preferences of a heterogeneous driver population.
The scope of global food technology market was estimated at USD 184.30 billion in 2023, USD 202.62 billion in 2024, and is projected to grow at a compound annual growth rate (CAGR) of 9.79% from 2024 to 2034, reaching approximately USD 515.83 billion. Technology is driving the growth of the food industry in various positive ways such as online food delivery in minutes, quality assessment, customer reviews, reducing hunger, and the like. But together with several advantages it also carries concerns like job displacement, food safety/security issues, regulatory compliance, and sustainability. To overcome these challenges, redesigning the digital food plate is critical in the form of concrete guidelines and regulations. Considering the above perspective, this chapter, adopting the analytical method, examines the role of digital and emerging technologies in shaping the food industry. Furthermore, it critically evaluates the way forward towards sustainability.
In reusable resources like vehicle sharing, city parking, and hotel services, resource units are used by consecutive customers for a stochastic usage time and are made available for future customers upon return. We model such reusable resources as Erlang loss systems, where customers who find all units occupied are blocked. Customers are rational decision-makers, and their willingness to join the system depends on factors such as price, time limits, service value, and resource availability. When each customer is charged a fixed price, we examine whether it is beneficial to install time limits, which increase availability at the expense of reduced usage time. We formulate and solve a two-dimensional revenue maximization problem to optimize both the usage price and the time limit. We differentiate between on-demand resources, which are immediately reusable upon return, and reserved resources, which can only be reused after the time limit expires. With on-demand resources, setting a time limit is proven to be detrimental, making the entry price the sole tool for maximizing revenue. This conclusion remains valid when the usage time increases with the entry price. In contrast, managing reserved resources involves an interplay between time limits and entry prices. Additionally, we show that loss systems (where customers do not wait) should be managed differently from systems where customers do wait. Specifically, in systems with waiting, we prove that only the time limit should be optimized, while the entry price should capture the full market share. We prove our results by analyzing the behavior of the equilibrium arrival rate and revenue, which necessitates establishing novel properties for the blocking probability. These properties yield new sharp bounds for the blocking probability and enable us to prove the convergence of the optimized model for on-demand resources to the QED regime as the number of resource units grows.
Practice- and Policy-Oriented Abstract
Mergers between on-demand service platforms have recently become prevalent, resulting in antitrust debates. On the one hand, mergers reduce market competition, raising antitrust concerns, such as potential increases in service prices. On the other hand, merged firms typically enable agents to serve all the customers from previously separate platforms, thereby gaining an advantage from resource pooling. To offer guidelines for resolving these debates, we investigate the impact of a merger between two on-demand service platforms. Our results show that because of the merger-enabled pooling benefit and a cross-side network effect (i.e., customers benefit from more agents and vice versa), a merger can result in a win-win-win outcome, where the platform profits, consumer surplus, and labor welfare are all improved after a merger. The win-win-win outcome is more probable if the premerger market is less saturated or if the market is more differentiated on the customer/agent side. Therefore, antitrust agencies might consider being more lenient toward a merger if the market has room to grow or if the two merging platforms have differentiated features.
A critical governance challenge for on-demand digital platforms is to increase the participation of their service providers. In this research, we design novel incentive structures by taking the unique features of on-demand digital platforms into account. In 12 micro randomized trials conducted in partnership with a major on-demand digital platform, we examine how combining monetary with nonmonetary incentives and providing them within a loss-aversion framework could motivate service providers to increase their participation levels. We show that in on-demand platforms the nonmonetary incentives inhibit the impact of monetary incentives on service provider participation once they are offered together. Furthermore, in contrast to traditional work settings, offering incentives within a loss-aversion framework only increases the effectiveness of nonmonetary incentives. We provide theoretical explanations and empirical examinations for these counterintuitive results. The insights from this research could be used by on-demand digital platforms to effectively mobilize and sustain their service providers’ participation to meet real-time stochastic demand.
This paper was accepted by Anindya Ghose, information systems.
Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.02603 .
Recently, ride‐sharing platforms have started to recruit not only with‐car drivers but also without‐car drivers. Without‐car drivers can directly rent cars from a ride‐sharing platform that operates its own fleets, which requires the platform to pay operating costs to manage the vehicles (self‐operated drivers) or rent cars from a car‐rental company that cooperates with the platform (rental drivers). This paper builds a stylized model to examine the impact of recruiting without‐car drivers on a platform's driver surplus, consumer surplus, and profit. We assume that a ride‐sharing platform first decides to recruit either with‐car drivers or both with‐ and without‐car drivers. Then, the platform sets a price charged to customers, a wage paid to drivers, and a rental fee charged to self‐operated drivers. Customers decide whether to order on a ride‐sharing platform based on the price and network externality term, while drivers choose to provide services according to the wage, rental fee charged by the platform or car‐rental company, and utilization rate. We find that when the platform recruits with‐car drivers (recruits with‐ and without‐car drivers), supply and demand are matched when the potential number of with‐car drivers is relatively small (large), while supply exceeds demand when the potential number of with‐car drivers is large (small). Moreover, when a ride‐sharing platform changes to recruit with‐ and without‐car drivers, the platform and customers always become better off, while drivers may not. Specifically, when the operating cost is relatively low or high (moderate), the drivers become better (worse) off.
Restaurants are increasingly relying on on-demand delivery platforms (e.g., DoorDash, Grubhub, and Uber Eats) to reach customers and fulfill takeout orders. Although on-demand delivery is a valuable option for consumers, whether restaurants benefit from or are being hurt by partnering with these platforms remains unclear. This paper investigates whether and to what extent the platform delivery channel substitutes restaurants’ own takeout/dine-in channels and the net impact on restaurant revenue. Empirical analyses show that restaurants overall benefit from on-demand delivery platforms—these platforms increase restaurants’ total takeout sales while creating positive spillovers to customer dine-in visits. However, the platform effects are substantially heterogeneous, depending on the type of restaurants (independent versus chain) and the type of customer channels (takeout versus dine-in). The overall positive effect on fast-food chains is four times as large as that on independent restaurants. For takeout, delivery platforms substitute independent restaurants’ but complement chain restaurants’ own takeout sales. For dine-in, delivery platforms increase both independent and chain restaurants’ dine-in visits by a similar magnitude. Therefore, the value of delivery platforms to independent restaurants mostly comes from the increase in dine-in visits, whereas the value to chain restaurants primarily comes from the gain in takeout sales. Further, the platform delivery channel facilitates price competition and reduces the opportunity for independent restaurants to differentiate with premium services and dine-in experience, which may explain why independent restaurants do not benefit as much from on-demand delivery platforms.
This paper was accepted by D. J. Wu, information systems.
Funding: Z. Li is grateful to the National Science Foundation Division of Social and Economic Sciences for support provided through the CAREER award [Grant 2243736].
Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2021.01010 .
“Quick commerce” refers to meeting consumers’ instant needs by delivering products ordered online within minutes. Although speed is thus inherently important, little is known about how deviations from communicated delivery times (whether late or early) might affect repurchase behavior. The authors study the effects of delivery time deviations on repurchase behavior using a large, customer-level, transaction data set from a Western European food delivery service and a controlled online experiment. The results show that late (early) deliveries increase (decrease) interpurchase times; these effects diminish with larger deviations. The results also show that late deliveries have a stronger effect on repurchase behavior than early deliveries of the same magnitude. The controlled online experiment establishes customer satisfaction as the underlying psychological mechanism that mediates the effect of delivery time deviations on repurchase behavior. These findings advance understanding of delivery time deviations and repurchase behavior by complementing prior research on disconfirmed waiting times and quick commerce. Practitioners can use the results to optimize their delivery algorithms, operations, and service recovery efforts.
On-demand meal delivery has become a feature of most cities around the world as a result of platforms and apps that facilitate it as well as the pandemic, which for a period, closed restaurants. Meals are delivered by couriers, typically on bikes, e-bikes, or scooters, who circulate collecting meals from kitchens and delivering them to customers, who usually order online. A Markov model for circulating couriers with n + 1 parameters, where [Formula: see text] is the number of kitchens plus customers, is derived by entropy maximization. There is one parameter for each kitchen and customer representing the demand for a courier, and there is one parameter representing the urgency of delivery. It is shown how the mean and variance of delivery time can be calculated once the parameters are known. The Markov model is irreducible. Two procedures are presented for calibrating model parameters on a data set of orders. Both procedures match known order frequencies with fitted visit probabilities; the first inputs an urgency parameter value and outputs mean delivery time, whereas the second inputs mean delivery time and outputs the corresponding urgency parameter value. Model calibration is demonstrated on a publicly available data set of meal orders from Grubhub. Grubhub data are also used to validate the calibrated model using a likelihood ratio. By changing the location of one kitchen, it is shown how the calibrated model can estimate the resulting change in demand for its meals and the corresponding mean delivery time. The Markov model could also be used for the assignment of courier trips to a street network.
History: This paper has been accepted for the Transportation Science Special Issue on ISTTT Conference.
Motivated by the diverse services provided by various on-demand service platforms to satisfy customers' different needs, we investigate the pricing of an on-demand crowdsourced delivery platform (CDP) with differentiated service levels. Using a multinomial logit model, we propose a two-sided market framework for investigating the pricing decisions of a CDP with two differentiated service levels under a general variable commission scheme and a fixed-rate commission scheme. The results show that (i) the two-sided pricing mechanism (price and wage decisions) allows the supply to match the demand for different service levels, except when the CDP adopts a fixed-rate commission scheme; (ii) depending on the magnitude of the fixed commission rate, adopting a fixed-rate commission scheme has different impacts on the equilibrium outcomes. We also extend our model by including the commission scheme and multiple service levels. The results show that (i) the CDP can achieve a higher optimal profit under the courier classification scheme than under a benchmark variable commission scheme, and (ii) offering senders and couriers diversified service levels benefits the CDP. Our work provides CDP practitioners with a quantitative tool for pricing for service-level menus.
Purpose
Although the impact of carbon emissions regulations is evident to upstream automakers, their influence on downstream B2C car-sharing platforms remains unclear. This article reveals the influence of carbon emission regulations on the performance of supply chain members. In particular, we focused on the decision of B2C car-sharing platforms.
Design/methodology/approach
We develop a three-stage dynamic game model consisting of an automaker, a B2C car-sharing platform and consumers.
Findings
The carbon emission cap has a critical threshold. Above this threshold, the regulation is ineffective for the platform’s operating model. Below it, the regulation affects the platform, moderated by customers' green awareness. The threshold initially decreases (weakly) and then increases in awareness. Effective caps reduce profits for the manufacturer, B2C car-sharing platform and supply chain, while ineffective caps see higher profits with increased awareness.
Originality/value
Firstly, this paper explores the impact of carbon emission caps on the operational strategies of B2C car-sharing platforms within the sharing economy, complementing existing research. Secondly, it identifies conditions where stricter caps prompt B2C car-sharing platforms to adjust their operational models and offers fresh insights for managers and departments responsible for carbon emission policy formulation. Thirdly, the study uncovers how carbon emission caps affect the performance of supply chain members, providing crucial managerial insights for sustainable operations.
Many third‐party platforms for shared manufacturing (platforms) have developed rapidly in recent years. The quality admission and pricing of production capacity are crucial issues for these platforms. This paper focuses on the platform supply chain composed of a platform, multiple manufacturers with surplus production capacity (sharers), and multiple manufacturers with insufficient production capacity (renters). Considering the impact of the production capacity quality on the shared scale and rental demand of production capacity, this paper constructs a game model for the platform supply chain under different quality admission scenarios, investigates the quality admission and price decisions of production capacity on the platform, and analyses the impact of the quality admission on the sharers’ and renters’ profits. The results show that the platform sets the quality admission when the sharers’ production capacity scale is small and the platform regulatory cost is low. The platform always increases its service price but may reduce the rental price of production capacity under the quality admission scenario. When the sharers’ production capacity scale is large, the platform lowers the threshold of quality admission with the quality elasticity of production capacity increasing. In addition, when the quality elasticity of production capacity and platform regulatory cost are low, and the sharers’ production capacity scale is small, the quality admission of the platform realizes the win–win–win situation for the platform, renters, and sharers.
Unexpected cancellation of food delivery orders poses significant challenges to resource allocation planning and could lead to reduced revenue for service providers. This paper addresses this issue by developing an optimization framework that can reassign canceled orders to alternative customers to maximize net revenue and minimize resource wastage. The problem is formulated as a route-based Markov decision process, named the Dynamic Routing and Pricing Problem with Cancellation (DRPPC). A solution approach based on the proximal policy optimization strategy is introduced as a computationally effective way of solving the optimization problem using reinforcement learning techniques. Experimental results demonstrate that the proposed computational method outperforms selected benchmark approaches with higher revenue from various kinds of scenarios with uncertainties. This research advances the intersection of urban logistics and reinforcement learning, offering actionable strategies for enhanced operational resilience in food delivery service providers.
Problem definition: Motivated by the debate around workers’ welfare in the gig economy, we propose a framework to evaluate current practices and possible alternatives. We study a setting in which customers seek service from workers and a platform facilitates such matches over the course of the day. The platform allocates time slots to workers using an allocation policy, and the workers are strategic agents (with respect to “when to work”) who maximize their expected utility that depends on their preferred times to work, the allocated slots, and the total availability time. The platform seeks to ensure that a sufficient number of workers is available to satisfy demand, whereas the workers aim to maximize their wage-driven utility. Methodology/results: We evaluate policies on two dimensions critical to any firm: the supply of workers across the day, and the effective wages of workers. We illustrate that several families of currently deployed policies have serious limitations. We find these limitations exist because the policies do not let workers fully express their preferences and/or cannot account for heterogeneity in such preferences. We propose a new allocation policy and establish strong performance guarantees with respect to both the workers’ supply and effective wages. The policy is simple and fully leverages the market information to reach better market outcomes. We supplement our theory with numerical experiments in the context of ride-hailing calibrated on various New York City data sets that illustrate performance across a range of markets. Managerial implications: We highlight a fundamental inefficiency of policies currently deployed that limit workers’ ability to express their preferences. By allowing workers to express their temporal preferences, and by judiciously prioritizing “full-time” workers over “part-time” workers, we can obtain a potentially significant Pareto improvement, maintaining (or even increasing) workers’ supply while increasing their effective wages.
Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.0414 .
One prevalent assumption in queueing theory is that the number of servers in a queueing model is deterministic. However, randomness in the number of available servers often arises in practice, e.g., in virtual call centers where agents are allowed to set their own schedules. In this paper, we study the problems of staffing and controlling queueing systems with an uncertain number of servers and impatient customers. Because randomness in the number of servers creates congestion in the system, the customer abandonment distribution plays an important role. We characterize how it affects both the optimal staffing policy and the cost incurred by the manager. Because of that strong dependence on the abandonment distribution, it is natural to investigate ways of controlling customer abandonment behavior so as to mitigate that cost. Here, we propose doing so by making delay announcements in the system. We characterise how the manager may use three controls in her toolbox, staffing, compensation, and the announcements, to effectively control her system. We show that despite jointly optimizing the usage of those three controls, it may be cost effective for the manager to understaff, overstaff, or match supply and demand in any given shift. This article is protected by copyright. All rights reserved.
Traditionally, research focusing on the design of routing and staffing policies for service systems has modeled servers as having fixed (possibly heterogeneous) service rates. However, service systems are generally staffed by people. Furthermore, people respond to workload incentives; that is, how hard a person works can depend both on how much work there is and how the work is divided between the people responsible for it. In a service system, the routing and staffing policies control such workload incentives; and so the rate servers work will be impacted by the system’s routing and staffing policies. This observation has consequences when modeling service system performance, and our objective in this paper is to investigate those consequences.
We do this in the context of the M/M/N queue, which is the canonical model for large service systems. First, we present a model for “strategic” servers that choose their service rate to maximize a trade-off between an “effort cost,” which captures the idea that servers exert more effort when working at a faster rate, and a “value of idleness,” which assumes that servers value having idle time. Next, we characterize the symmetric Nash equilibrium service rate under any routing policy that routes based on the server idle time (such as the longest idle server first policy). We find that the system must operate in a quality-driven regime, in which servers have idle time, for an equilibrium to exist. The implication is that to have an equilibrium solution the staffing must have a first-order term that strictly exceeds that of the common square-root staffing policy. Then, within the class of policies that admit an equilibrium, we (asymptotically) solve the problem of minimizing the total cost, when there are linear staffing costs and linear waiting costs. Finally, we end by exploring the question of whether routing policies that are based on the service rate, instead of the server idle time, can improve system performance.
Abstract Any buyer that depends,on suppliers for the delivery of a service or the production,of a make-to-order component,should pay close attention to the suppliers’ service or delivery lead times. This paper studies a queuing,model in which,two strategic servers choose their capacities/processing rates and faster service is costly. The buyer allocates demand,to the servers based on their performance; the faster a server works, the more demand the server is allocated. The buyer’s objective is to minimize,the average,lead time received from the servers. There are two important,attributes to consider in the design of an allocation policy: the degree to which the allocation policy effectively utilizes the servers’ capacities and the strength of the incentives the allocation policy provides for the servers to work,quickly. Previous research suggests that there exists a tradeoff between efficiency and incentives, i.e., in the choice between two allocation policies a buyer may prefer the less efficient one because,it provides stronger incentives. We find considerable variation in the performance,of allocation policies: some,intuitively rea-
Internet-enabled markets are becoming viable venues for procurement of professional services. We investigate bidding behavior within the most active area of these early knowledge markets---the market for software development. These markets are important both because they provide an early view of the effectiveness of online service markets and because they have a potentially large impact on how software development services are procured and provided. Using auction theory, we develop a theoretical model that relates market characteristics to bidding and transaction behavior, taking into account costly bidding. We then test our model using data from an active online market for software development services, which yields contracts for 30%--40% of posted projects. In its current format, however, the studied market may induce excessive bidding by vendors. Consistent with our theoretical predictions and those of Carr (2003), higher-value projects attract significantly more bids, with lower average quality. Greater numbers of bids raise the cost to all participants, due to costly bidding and bid evaluation. Perhaps as a consequence, higher-value projects are also much less likely to be awarded.
This paper considers a profit-maximizing make-to-order manufacturer that offers multiple products to a market of price and delay sensitive users, using a model that captures three aspects of particular interest: first, the joint use of dynamic pricing and lead-time quotation controls to manage demand; second, the presence of a dual sourcing mode that can expedite orders at a cost; and third, the interaction of the aforementioned demand controls with the operational decisions of sequencing and expediting that the firm must employ to optimize revenues and satisfy the quoted lead times. Using an approximating diffusion control problem we derive near-optimal dynamic pricing, lead-time quotation, sequencing, and expediting policies that provide structural insights and lead to practically implementable recommendations. A set of numerical results illustrates the value of joint pricing and lead-time control policies.
This paper considers pricing and capacity sizing decisions, in a single-class Markovian model motivated by communication and information services. The service provider is assumed to operate a finite set of processing resources that can besharedamong users; however, this shared mode of operation results in a service-rate degradation. Users, in turn, are sensitive to the delay implied by the potential degradation in service rate, and to the usage fee charged for accessing the system. We study the equilibrium behavior of such systems in the specific context of pricing and capacity sizing under revenue and social optimization objectives. Exact solutions to these problems can only be obtained via exhaustive simulations. In contrast, we pursue approximate solutions that exploit large-capacity asymptotics. Economic considerations and natural scaling relations demonstrate that the optimal operational mode for the system is close to "heavy traffic." This, in turn, supports the derivation of simple approximate solutions to economic optimization problems, via asymptotic methods that completely alleviate the need for simulation. These approximations seem to be extremely accurate. The main insights that are gleaned in the analysis follow: congestion costs are "small," the optimal price admits a two-part decomposition, and the joint capacity sizing and pricing problem decouples and admits simple analytical solutions that are asymptotically optimal. All of the above phenomena are intimately related to statistical economies of scale that are an intrinsic part of these systems.
Should a firm charge on a per-use basis or sell subscriptions when its service experiences congestion? Queueing-based models of pricing primarily focus on charging a fee per use for the service, in part because per-use pricing enables the firm to regulate congestion--raising the per-use price naturally reduces how frequently customers use a service. The firm has less control over usage with subscription pricing (by definition, with subscription pricing customers are not charged proportional to their actual usage), and this is a disadvantage when customers dislike congestion. However, we show that subscription pricing is more effective at earning revenue. Consequently, the firm may be better off with subscription pricing, even, surprisingly, when congestion is intuitively most problematic for the firm: e.g., as congestion becomes more disliked by consumers. We show that the absolute advantage of subscription pricing relative to per-use pricing can be substantial, whereas the potential advantage of per-use pricing is generally modest. Subscription pricing becomes relatively more attractive if consumers become more heterogeneous in their service rates (e.g., some know they are "heavy" users and others know they are "light" users) as long as capacity is fixed, the potential utilization is high, and the two segments have substantially different usage rates. Otherwise, heterogeneity in usage rates makes subscription pricing less attractive relative to per-use pricing. We conclude that subscription pricing can be effective even if congestion is relevant for the overall quality of a service.
We study a profit-maximizing firm providing a service to price and delay sensitive customers. We are interested in analyzing the scale economies inherent in such a system. In particular, we study how the firm's pricing and capacity decisions change as the scale, measured by the potential market for the service, increases. These decisions turn out to depend intricately on the form of the delay costs seen by the customers; we characterize these decisions up to the dominant order in the scale for both convex and concave delay costs. We show that when serving customers on a first-come, first-served basis, if the customers' delay costs are strictly convex, the firm can increase its utilization and extract profits beyond what it can do when customers' delay costs are linear. However, with concave delay costs, the firm is forced to decrease its utilization and makes less profit than in the linear case. While studying concave delay costs, we demonstrate that these decisions depend on the scheduling policy employed as well. We show that employing the last-come, first-served rule in the concave case results in utilization and profit similar to the linear case, regardless of the actual form of the delay costs.
It takes time to process purchases and as a result a queue of customers may form. The pricing and service rate decissions of a monopolist who must take this into account are characterized. We find that an increase in the average number of customers arriving in the market either has no effect on the monopoly price, or else causes the monopolist to reduce the price in the short run. In the long run the monopolist will increase the service rate and raise the price. When customer preferences are linear the equilibrium is socially efficient. When preferences are not linear equilibrium will not normally be socially efficient.
Problem definition: We consider an intermediary’s problem of dynamically matching demand and supply of heterogeneous types in a periodic-review fashion. Specifically, there are two disjoint sets of demand and supply types, and a reward for each possible matching of a demand type and a supply type. In each period, demand and supply of various types arrive in random quantities. The platform decides on the optimal matching policy to maximize the expected total discounted rewards, given that unmatched demand and supply may incur waiting or holding costs, and will be fully or partially carried over to the next period. Academic/practical relevance: The problem is crucial to many intermediaries who manage matchings centrally in a sharing economy. Methodology: We formulate the problem as a dynamic program. We explore the structural properties of the optimal policy and propose heuristic policies. Results: We provide sufficient conditions on matching rewards such that the optimal matching policy follows a priority hierarchy among possible matching pairs. We show that those conditions are satisfied by vertically and unidirectionally horizontally differentiated types, for which quality and distance determine priority, respectively. Managerial implications: The priority property simplifies the matching decision within a period, and the trade-off reduces to a choice between matching in the current period and that in the future. Then the optimal matching policy has a match-down-to structure when considering a specific pair of demand and supply types in the priority hierarchy.
Problem definition: We study the following puzzle: why do firms sometimes nurture long wait lines, rather than raising prices (to eliminate these lines)? Academic/practical relevance: In economic theory, prices are studied as a signal of quality, in absence of congestion (wait lines). We study price signaling when customers not only infer quality information from the price, but also from wait lines. Methodology: We developed a stylized model based on queuing and price signaling theory. The firm, privately informed about its quality (high or low), first determines its price. Next, customers arrive according to a Poisson process, observe the price and the wait line upon arrival, and decide whether to purchase. Results: When the proportion of informed consumers is low, to separate on price a high-quality firm must raise its price above the monopoly price. We show that there exist pooling equilibria in which firms instead charge a price lower than the monopoly price. We characterize two pooling equilibria. In both equilibria, a high-quality firm on average has a longer queue than a low-quality one, and long lines signal high quality. In the first one, the price is sufficiently low that short lines in turn signal low quality, so that the queue length almost perfectly reveals the type of the firm. In the second one, the price is in an intermediate range (but remains lower than the monopoly price), and short lines are an imperfect signal of quality. In each of the pooling equilibria, the high-quality firm earns a higher profit than in the separating equilibrium, despite the longer lines. Managerial implications: An empirical implication of our model is that, over time, when more customers become informed about the quality, the price of a new high-quality good should increase and the expected waiting lines should decrease. Our model integrates elements of price signaling in economics and equilibrium queueing in operations research. We hope that our research stimulates further interest in signaling games in service operations environments.
Understand the Strategic Behavior in Queueing Systems Rational Queueing provides one of the first unified accounts of the dynamic aspects involved in the strategic behavior in queues. It explores the performance of queueing systems where multiple agents, such as customers, servers, and central managers, all act but often in a noncooperative manner. The book first addresses observable queues and models that assume state-dependent behavior. It then discusses other types of information in queueing systems and compares observable and unobservable variations and incentives for information disclosure. The next several chapters present relevant models for the maximization of individual utilities, social welfare, and profits. After covering queueing networks, from simple parallel servers to general network structures, the author describes models for planned vacations and forced vacations (such as breakdowns). Focusing on supply chain models, he then shows how agents of these models may have different goals yet they all profit when the system operates efficiently. The final chapter allows bounded rationality by lowering the assumption of fully rational agents.
We describe an equilibrium model of peer-to-peer product sharing, or collaborative consumption, where individuals with varying usage levels make decisions about whether or not to own a homogeneous product. Owners are able to generate income from renting their products to nonowners while nonowners are able to access these products through renting on an as-needed basis. We characterize equilibrium outcomes, including ownership and usage levels, consumer surplus, and social welfare. We compare each outcome in systems with and without collaborative consumption and examine the impact of various problem parameters. Our findings indicate that collaborative consumption can result in either lower or higher ownership and usage levels, with higher ownership and usage levels more likely when the cost of ownership is high. Our findings also indicate that consumers always benefit from collaborative consumption, with individuals who, in the absence of collaborative consumption, are indifferent between owning and not owning benefitting the most. We study both profit-maximizing and social-welfare–maximizing platforms and compare equilibrium outcomes under both in terms of ownership, usage, and social welfare. We find that the difference in social welfare between the profit-maximizing and social-welfare–maximizing platforms is relatively modest.
The online appendix is available at https://doi.org/10.1287/mnsc.2017.2970 .
This paper was accepted by Gad Allon, operations management.
We analyze a service firm that caters to price and delay sensitive customers who are differentiated on both their value for the service and the cost of waiting. There is a continuum of customer types in our setting and we model each customer's cost of waiting to be linear in the delay incurred with a multiplier that is an increasing linear or sub-linear function of the customer's value for the service. Using a large system approach, we characterize the firm's revenue maximizing menu of price and delay quotations and the value of customer differentiation. We further characterize the value of offering coarse or few service grades and find that offering two service grades is asymptotically optimal on the typical square-root scale, relative to the optimal policy.
Recent platforms, like Uber and Lyft, offer service to consumers via “self-scheduling” providers who decide for themselves how often to work. These platforms may charge consumers prices and pay providers wages that both adjust based on prevailing demand conditions. For example, Uber uses a “surge pricing” policy, which pays providers a fixed commission of its dynamic price. With a stylized model that yields analytical and numerical results, we study several pricing schemes that could be implemented on a service platform, including surge pricing. We find that the optimal contract substantially increases the platform’s profit relative to contracts that have a fixed price or fixed wage (or both), and although surge pricing is not optimal, it generally achieves nearly the optimal profit. Despite its merits for the platform, surge pricing has been criticized because of concerns for the welfare of providers and consumers. In our model, as labor becomes more expensive, providers and consumers are better off with surge pricing because providers are better utilized and consumers benefit both from lower prices during normal demand and expanded access to service during peak demand. We conclude, in contrast to popular criticism, that all stakeholders can benefit from the use of surge pricing on a platform with self-scheduling capacity.
The e-companion is available at https://doi.org/10.1287/msom.2017.0618.
To investigate whether peer-to-peer rental markets for durable goods are welfare-improving, we develop a new dynamic model of such markets in which users with heterogeneous utilization rates may also trade in secondary markets. We calibrate our model with US automobile industry data and transaction-level data from Getaround, a large peer-to-peer car rental marketplace. Counterfactual analyses illustrate significant shifts away from asset ownership as marketplace access grows. Used-good prices fall and replacement rates rise, while gains in consumer surplus range from 0.8% to 6.6%. The changes in consumption mix and the surplus increases are significantly more pronounced for below-median income consumers.
Participants in matching markets face search and screening costs which prevent the market from clearing efficiently. In many settings, the rise of online matching platforms has dramatically reduced the cost of finding and contacting potential partners. While one might expect both sides of the market to benefit from reduced search costs, this is far from guaranteed. In particular, this change may force participants to screen more potential partners before finding one who is willing to accept their offer. Thus, for a fixed screening cost, reductions in search and application costs may actually decrease aggregate welfare.We illustrate this fact using a model in which agents apply and screen asynchronously. In our model, it is possible that an agent on one side of the market (an 'employer') identifies a suitable match on the other side (an 'applicant'), only to find that this applicant has already matched. We find that equilibrium is generically inefficient for both employers and applicants. Most notably, when application costs are sufficiently small, uncertainty about applicant availability may drive equilibrium employer welfare to zero.We consider a simple intervention available to the platform: limiting the visibility of applicants. We find that this intervention can significantly improve the welfare of agents on both sides of the market; applicants pay lower application costs, while employers are less likely to find that the applicants they screen have already matched. Somewhat counterintuitively, the benefits of showing fewer applicants to each employer are greatest in markets in which there is a shortage of applicants.
We study a multiserver queueing model of a revenue-maximizing firm providing a service to a market of heterogeneous price- and delay-sensitive customers with private individual preferences. The firm may offer a selection of service classes that are differentiated in prices and delays. Using a deterministic relaxation, which simplifies the problem by preserving the economic aspects of price-and-delay differentiation while ignoring queueing delays, we construct a solution to the fully stochastic problem that is incentive compatible and near optimal in systems with large capacity and market potential. Our approach provides several new insights for large-scale systems: (i) the deterministic analysis captures all pricing, differentiation, and delay characteristics of the stochastic solution that are nonnegligible at large scale; (ii) service differentiation is optimal when the less delay-sensitive market segment is sufficiently elastic; (iii) the use of “strategic delay” depends on system capacity and market heterogeneity—and it contributes significant delay when the system capacity is underutilized or when the firm offers three or more service classes; and (iv) connecting economic optimization to queueing theory, the revenue-optimized system has the premium class operating in a “quality-driven” regime and the lower-tier service classes operating with noticeable delays that arise either endogenously (“efficiency-driven” regime) or with the addition of strategic delay by the service provider.
This paper was accepted by Gérard Cachon, stochastic models and simulation.
Peer-to-peer markets, collectively known as the sharing economy, have emerged as alternative suppliers of goods and services traditionally provided by long-established industries. The authors explore the economic impact of the sharing economy on incumbent firms by studying the case of Airbnb, a prominent platform for short-term accommodations. They analyze Airbnb's entry into the state of Texas and quantify its impact on the Texas hotel industry over the subsequent decade. In Austin, where Airbnb supply is highest, the causal impact on hotel revenue is in the 8%-10% range; moreover, the impact is nonuniform, with lower-priced hotels and hotels that do not cater to business travelers being the most affected. The impact manifests itself primarily through less aggressive hotel room pricing, benefiting all consumers, not just participants in the sharing economy. The price response is especially pronounced during periods of peak demand, such as during the South by Southwest festival, and is due to a differentiating feature of peer-to-peer platforms-enabling instantaneous supply to scale to meet demand.
Peer-to-peer markets such as eBay, Uber, and Airbnb allow small suppliers to compete with traditional providers of goods or services. We view the primary function of these markets as making it easy for buyers to find sellers and engage in convenient, trustworthy transactions. We discuss elements of market design that make this possible, including search and matching algorithms, pricing, and reputation systems. We then develop a simple model of how these markets enable entry by small or flexible suppliers, and how they impact existing firms. Finally, we consider the regulation of peer-to-peer markets and the economic arguments for different approaches to licensing and certification, data, and employment regulation.
Recent technological advances in online and mobile communications have enabled collaborative consumption or product sharing among consumers on a massive scale. Collaborative consumption has emerged as a major trend as the global economic recession and social concerns about consumption sustainability lead consumers and society as a whole to explore more efficient use of resources and products. We develop an analytical framework to examine the strategic and economic impact of product sharing among consumers. A consumer who purchased a firm’s product can derive different usage values across different usage periods. In a period with low self-use value, the consumer may generate some income by renting out her purchased product through a third-party sharing platform as long as the rental fee net of transaction costs exceeds her own self-use value. Our analysis shows that transaction costs in the sharing market have a nonmonotonic effect on the firm’s profits, consumer surplus, and social welfare. We find that when the firm strategically chooses its retail price, consumers’ sharing of products with high marginal costs is a win-win situation for the firm and the consumers, whereas their sharing of products with low marginal costs can be a lose-lose situation. Furthermore, in the presence of the sharing market, the firm will find it optimal to strategically increase its quality, leading to higher profits but lower consumer surplus.
This paper was accepted by J. Miguel Villas-Boas, marketing.
How should a firm design a price/lead-time menu and scheduling policy to maximize revenues from heterogeneous time-sensitive customers with private information about their preferences? We consider a queueing system with multiple customer types that differ in their valuations for instant delivery and their delay costs. The distinctive feature of our model is that the ranking of customer preferences depends on lead times: patient customers are willing to pay more than impatient customers for long lead times, and vice versa for speedier service. We provide necessary and sufficient conditions, in terms of the capacity, the market size, and the properties of the valuation-delay cost distribution, for three features of the optimal menu and segmentation: pricing out the middle of the delay cost spectrum while serving both ends, pooling types with different delay costs into a single class, and strategic delay to deliberately inflate lead times.
This study illustrates how a manufacturer can use leadtime differentiation-selling the same product to different customers at different prices based on delivery leadtime-to simultaneously increase revenue and reduce capacity requirements. The manufacturer's production facility is modeled as an exponential single-server queue with two classes of customers that differ in price sensitivity and delay sensitivity. The manufacturer chooses the service rate and a static price for each class of customer, and then dynamically quotes leadtimes to potential customers and decides the order in which customers are processed. The arrival rate for each class decreases linearly with price and leadtime. The manufacturer's objective is to maximize profit, subject to the constraint that each customer must be processed within the promised leadtime. Assuming that some customers will tolerate a long delivery leadtime, we show that this problem has a simple near-optimal solution. Under our proposed policy, capacity utilization is near 100%. Impatient customers pay a premium for immediate delivery and receive priority in scheduling, whereas patient customers are quoted a leadtime proportional to the current queue length. Queue length and leadtime can be closely approximated by a reflected Ornstein-Uhlenbeck diffusion process. Hence, we have a closed form expression for profit, and choose prices and capacity to optimize this. In case customers may choose either the class I deal or the class 2 deal, the proposed policy is made incentive compatible by quoting a leadtime for the class 2 (patient) customers that is longer than the actual queueing delay.
Online service marketplaces allow service buyers to post their project requests and service providers to bid for them. To reduce the transactional risks, marketplaces typically track and publish previous seller performance. By analyzing a detailed transactional data set with more than 1,800,000 bids corresponding to 270,000 projects posted between 2001 and 2010 in a leading online intermediary for software development services, we empirically study the effects of the reputation system on market outcomes. We consider both a structured measure summarized in a numerical reputation score and an unstructured measure based on the verbal praise left by previous buyers, which we encode using text mining techniques. We find that buyers trade off reputation (both structured and unstructured) and price and are willing to accept higher bids posted by more reputable bidders. Sellers also respond to changes in their own reputation through three different channels. They increase their bids with their reputation score (price effect) but primarily use a superior reputation to increase their probability of being selected (volume effect) as opposed to increasing their bid prices. Negative shocks in seller reputation are associated to an increase in the probability of seller exit (exit effect), but this effect is moderated by the investment that the seller has made in the site. We conclude that participants in this market are very responsive to the numerical reputation score and also to the unstructured reputational information, which behaves in a similar way to the structured numerical reputation score but provides complementary information.
Preface. 1. Introduction. 2. Observable Queues. 3. Unobservable Queues. 4. Priorities. 5. Reneging and Jockeying. 6. Schedules and Retrials. 7. Competition Among Servers. 8. Service Rate Decisions. Index.
Consider a system that is modeled as an M/M/1 queueing system with multiple user classes. Each class is characterized by its delay cost per unit of time, its expected service time and its demand function. This paper derives a pricing mechanism which is optimal and incentive- compatible in the sense that the arrival rates and execution priorities jointly maximize the expected net value of the system while being determined, on a decentralized basis, by individual users. A closed-form expression for the resulting price structure is presented and studied.
This chapter reviews the supply chain coordination with contracts. Numerous supply chain models are discussed. In each model, the supply chain optimal actions are identified. The chapter extends the newsvendor model by allowing the retailer to choose the retail price in addition to the stocking quantity. Coordination is more complex in this setting because the incentives provided to align one action might cause distortions with the other action. The newsvendor model is also extended by allowing the retailer to exert costly effort to increase demand. Coordination is challenging because the retailer's effort is noncontractible—that is, the firms cannot write contracts based on the effort chosen. The chapter also discusses an infinite horizon stochastic demand model in which the retailer receives replenishments from a supplier after a constant lead time. Coordination requires that the retailer chooses a large basestock level.
How should a capacity-constrained firm design an incentive-compatible price-scheduling mecha-nism to maximize revenues from a heterogeneous pool of time-sensitive customers with private information on their willingness to pay, time-sensitivity and processing requirement? We con-sider this question in the context of a queueing system that serves two customer types. We provide the following insights. First, the familiar cµ priority rule, known to minimize the system-wide expected delay cost and to be incentive-compatible under social optimization, need not be optimal in this setting. This specific fact suggests a more general guideline: in design-ing incentive-compatible and revenue-maximizing scheduling policies, delay cost-minimization, which plays a prominent role in controlling and pricing queueing systems, should not be the dominant criterion ex ante. Second, we identify optimal scheduling policies with novel features. One such policy prioritizes the more time-sensitive customers but voluntarily delays the com-pleted orders of low-priority customers. This insertion of strategic delay deters time-sensitive customers from purchasing the low-priority class. In other situations, it is optimal to appropri-ately randomize priority assignments, in one extreme case serving customers in the reverse cµ order, which maximizes the system delay cost among all work conserving policies. Compared to the cµ rule, these optimal policies increase, decrease or reverse the delay differentiation between customer types. We show how the optimal level of delay differentiation systematically emerges from a trade-off between operational constraints and customer incentives. Third, our stepwise solution approach can be adapted for designing revenue-maximizing and incentive-compatible mechanisms in systems with different customer attributes or operational properties.
We consider a make-to-order …rm that has the ability to dynamically o¤er menus of prices and production (or service) leadtimes to its customers. Customers seeking a particular product (or service) will choose from the o¤ered menu the pair of prices and leadtimes that maximizes their value for the product minus their delay cost and the price for that leadtime. Dynamic control allows the retailer to tune leadtimes and prices to the current backlog. We consider two classes of customers who have the same valuation for the product but di¤er in their level of patience (a concept made precise in the paper). We investigate how such dynamic menus should be chosen in the context of a large capacity asymptotic regime and propose policies when customers'leadtime costs are convex-concave. We consider both the full information case and the (more realistic) case where the …rm, being unaware of customer type, must o¤er incentive-compatible menus. We propose readily-implementable policies and test them numerically against a number of natural benchmarks.
In many services, the quality or value provided by the service increases with the time the service provider spends with the customer. However, longer service times also result in longer waits for customers. We term such services, in which the interaction between quality and speed is critical, as customer-intensive services. In a queueing framework, we parameterize the degree of customer intensity of the service. The service speed chosen by the service provider affects the quality of the service through its customer intensity. Customers queue for the service based on service quality, delay costs, and price. We study how a service provider facing such customers makes the optimal “quality–speed trade-off.” Our results demonstrate that the customer intensity of the service is a critical driver of equilibrium price, service speed, demand, congestion in queues, and service provider revenues. Customer intensity leads to outcomes very different from those of traditional models of service rate competition. For instance, as the number of competing servers increases, the price increases, and the servers become slower.
This paper was accepted by Sampath Rajagopalan, operations and supply chain management.
In this paper, we study a service network in which an agency is responsible for satisfying a constraint on the expected waiting and service time experienced by customers. However, the agency does not render the actual service. Instead, it serves to coordinate independently operated facilities. The coordinating agency must devise a strategy for allocating compensation and customers to the self-interested operators in order to minimize its own costs. For a network of two facilities, we model the facilities' self-interested capacity decisions as the solution to a game. Using this analytical framework, we compare two types of customer allocation: one from a common queue, and one from separate queues. Our analysis shows that it can be in the best interest of the coordinating agency to adopt a separate queue allocation scheme instead of one based on a common queue. Although doing so sacrifices risk-pooling benefits, these can be more than offset by the stronger incentives that are created for the independent facilities.
This article studies the effects of queueing delays, and users' related costs, on the management and control of computing resources. It offers a methodology for setting price, utilization, and capacity, taking into account the value of users' time, and it examines the implications of alternative control structures, determined by the financial responsibility assigned to the data processing manager.
A queueing model--together with a cost structure--is presented, which envisages the imposition of tolls on newly arriving customers. It is shown that frequently this is a strategy which might lead to the attainment of social optimality.
Designing promotions to scale marketplaces. Working paper, INSEAD
- A Kabra
- E Belavina
- K Girotra
Joes: Agent pricing behavior in the sharing economy. Working paper
- J Li
- A Moreno
- D J Zhang
Incentive based service system design: Staffing and compensation to trade off speed and quality. Working paper
- D Zhan
- A Ward