Ride-hailing platforms offer riders pooling service to share rides with other riders. On the one hand, pooling service mitigates congestion and decreases rider wait times in rush hours, and sharing riders benefit from reduced prices. On the other hand, sharing riders may compromise on privacy and space when riding with strangers, and may take more time to reach their destinations. We derive a queueing model that considers solo ride and shared ride together, where riders are strategic in choosing which ride to participate. We analyze and compare the decentralized rider decisions and the centralized social planner decisions. In most cases, a smaller fraction of riders choose shared rides compared to that under the socially optimal decision, which we call under-share. Nonetheless, riders can always be induced to choose the socially optimal strategy in equilibrium under a proper monetary, social, or priority scheme. Interestingly, under the priority scheme, overshare can happen, whereas without the priority scheme the decentralized riders behave social-optimally in the same arrival rate and sharing externality range. In contrast to that individual riders always over-join an unobservable M/M/1 queue (compared to the social optimal), riders always under-join the queue in our model with an additional ride-sharing option. Moreover, the social planner may restrict the number of riders even if the arrival rate is below the socially optimal one. This is because social welfare may be a bimodal function of the arrival rate under the assumption that all riders join. At last, we conduct a numerical study with the ride-hailing data of Chicago, and discover that though under-share occurs in residential areas during morning rush hours and in downtown during evening rush hours, the observed sharing fractions are very close to the optimal ones. Over-share occurs during the same time interval but in the opposite areas.
{"title":"Share or Solo? Individual and Social Choices in Ride-Hailing","authors":"Ming Hu, Jianfu Wang, Hengda Wen","doi":"10.2139/ssrn.3675050","DOIUrl":"https://doi.org/10.2139/ssrn.3675050","url":null,"abstract":"Ride-hailing platforms offer riders pooling service to share rides with other riders. On the one hand, pooling service mitigates congestion and decreases rider wait times in rush hours, and sharing riders benefit from reduced prices. On the other hand, sharing riders may compromise on privacy and space when riding with strangers, and may take more time to reach their destinations. We derive a queueing model that considers solo ride and shared ride together, where riders are strategic in choosing which ride to participate. We analyze and compare the decentralized rider decisions and the centralized social planner decisions. In most cases, a smaller fraction of riders choose shared rides compared to that under the socially optimal decision, which we call under-share. Nonetheless, riders can always be induced to choose the socially optimal strategy in equilibrium under a proper monetary, social, or priority scheme. Interestingly, under the priority scheme, overshare can happen, whereas without the priority scheme the decentralized riders behave social-optimally in the same arrival rate and sharing externality range. In contrast to that individual riders always over-join an unobservable M/M/1 queue (compared to the social optimal), riders always under-join the queue in our model with an additional ride-sharing option. Moreover, the social planner may restrict the number of riders even if the arrival rate is below the socially optimal one. This is because social welfare may be a bimodal function of the arrival rate under the assumption that all riders join. At last, we conduct a numerical study with the ride-hailing data of Chicago, and discover that though under-share occurs in residential areas during morning rush hours and in downtown during evening rush hours, the observed sharing fractions are very close to the optimal ones. Over-share occurs during the same time interval but in the opposite areas.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"63 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85603622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract We study capacity decisions in a service environment with seasonal arrival rates and strategic customers. Due to seasonality, the firm uses a postponement strategy where compensation is given to the peak-time customers who cannot be served in the peak time and agree to delay service. We extend the analysis of demand postponement to the case with strategic customers, i.e., off-peak customers who may pretend to be peak-time customers purely to receive the postponement discount. Focusing on the capacity and discount decisions for the service provider, we develop a game-theoretic model to study when the service provider should consider using a postponement strategy in the presence of strategic customers. At the first stage, the service provider decides on the capacity level and discount. At the second stage, strategic customers decide whether to show up in the peak time. We find that strategic customers exhibit a “follow-the-crowd” type of behavior. Unsurprisingly, we find that the presence of strategic customers does not benefit the service provider. More interestingly, we find that the presence of strategic customers may benefit the peak-time customers, unlike the inventory literature with strategic (arrival-timing) customers, in which the supplier’s response is usually detrimental to non-strategic customers. In the presence of strategic customers, the supplier needs to offer a lower discount or build more capacity than without strategic customers. If the supplier does not have control over the discount value, she might need to build excess capacity to prevent the strategic customers from showing up in the peak time.
{"title":"Demand Postponement with Strategic Service Customers","authors":"Haiyan Wang, T. Olsen, Timofey Shalpegin","doi":"10.2139/ssrn.3671121","DOIUrl":"https://doi.org/10.2139/ssrn.3671121","url":null,"abstract":"Abstract We study capacity decisions in a service environment with seasonal arrival rates and strategic customers. Due to seasonality, the firm uses a postponement strategy where compensation is given to the peak-time customers who cannot be served in the peak time and agree to delay service. We extend the analysis of demand postponement to the case with strategic customers, i.e., off-peak customers who may pretend to be peak-time customers purely to receive the postponement discount. Focusing on the capacity and discount decisions for the service provider, we develop a game-theoretic model to study when the service provider should consider using a postponement strategy in the presence of strategic customers. At the first stage, the service provider decides on the capacity level and discount. At the second stage, strategic customers decide whether to show up in the peak time. We find that strategic customers exhibit a “follow-the-crowd” type of behavior. Unsurprisingly, we find that the presence of strategic customers does not benefit the service provider. More interestingly, we find that the presence of strategic customers may benefit the peak-time customers, unlike the inventory literature with strategic (arrival-timing) customers, in which the supplier’s response is usually detrimental to non-strategic customers. In the presence of strategic customers, the supplier needs to offer a lower discount or build more capacity than without strategic customers. If the supplier does not have control over the discount value, she might need to build excess capacity to prevent the strategic customers from showing up in the peak time.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"113 1-2 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76892482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper examines a development process in which a firm explores alternative approaches for the development of a single product. To do this, the firm employs competing internal teams, but it also needs to involve key suppliers who are responsible for the development of a key component (or subsystem) via a collaboration with one or several internal teams. We explore how internal competition affects the suppliers' collaborative efforts and the firm's profits, and also how the firm should allocate the internal teams to potential suppliers. Our analysis suggests that this setting is significantly different from standard tournament settings, implying that existing research results should not be simply extrapolated to supply chain settings. For example, unlike in traditional tournaments, we find that increasing supplier competition by increasing the number of teams per supplier can actually increase supplier efforts. We develop a tournament model with two suppliers who differ both in their expertise for the specific type of development project and in their development costs (i.e., their cost-effectiveness). We also identify a critical influential factor: the cost savings resulting from the similarity between different development efforts by the same supplier (with different internally competing teams). If the cost savings are not very high, i.e., if the different development efforts are not very similar, the firm should optimally allocate more teams to the supplier with less relevant expertise or lower cost-effectiveness. Similarly, involving a third supplier is beneficial only if the cost savings are low. While companies make significant efforts to identify the supplier with the best expertise, our results show that allocating more teams to the stronger supplier generally undermines the suppliers' incentives to exert high levels of effort. Hence, allocating the most teams to that supplier might not always improve firm profits.
{"title":"Supplier Collaboration in Collaborative Product Development with Internal Competition","authors":"Timofey Shalpegin, Svenja C. Sommer, C. van Delft","doi":"10.2139/ssrn.3671975","DOIUrl":"https://doi.org/10.2139/ssrn.3671975","url":null,"abstract":"This paper examines a development process in which a firm explores alternative approaches for the development of a single product. To do this, the firm employs competing internal teams, but it also needs to involve key suppliers who are responsible for the development of a key component (or subsystem) via a collaboration with one or several internal teams. We explore how internal competition affects the suppliers' collaborative efforts and the firm's profits, and also how the firm should allocate the internal teams to potential suppliers. Our analysis suggests that this setting is significantly different from standard tournament settings, implying that existing research results should not be simply extrapolated to supply chain settings. For example, unlike in traditional tournaments, we find that increasing supplier competition by increasing the number of teams per supplier can actually increase supplier efforts. We develop a tournament model with two suppliers who differ both in their expertise for the specific type of development project and in their development costs (i.e., their cost-effectiveness). We also identify a critical influential factor: the cost savings resulting from the similarity between different development efforts by the same supplier (with different internally competing teams). If the cost savings are not very high, i.e., if the different development efforts are not very similar, the firm should optimally allocate more teams to the supplier with less relevant expertise or lower cost-effectiveness. Similarly, involving a third supplier is beneficial only if the cost savings are low. While companies make significant efforts to identify the supplier with the best expertise, our results show that allocating more teams to the stronger supplier generally undermines the suppliers' incentives to exert high levels of effort. Hence, allocating the most teams to that supplier might not always improve firm profits.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"52 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83093961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Purpose: This paper aims to participate in answering the following research question: Will blockchain become the norm in supply chain management? Valuable insights from academia and industry about the use of blockchain technology in supply chain management are presented.
Design/methodology/approach: I review the literature that mainly examined the use of blockchain technology in supply chain management from 2008 to 2020. Additionally, I explore two blockchain projects: Nimble and Carrefour.
Findings: The mainstream research on the use of blockchain technology in supply chain management exhibits a positive trend. This study shows that relevant academic research increased by 174% in only 11 months (from February 2018 to January 2019). Countries and supply chains, to the same extent, are making bold moves toward using blockchain on a large scale. The European Commission, for example, funded a project to create a federated, multi-sided business ecosystem based on cloud services. The project developed a platform, named Nimble, to enable supply chain logistics to benefit from internet platforms and blockchain services. In another example, Carrefour has implemented blockchain in nine sectors (chicken, tomatoes, eggs, oranges, etc.) with a target of 300 sectors by 2022.
Originality/value: Although this area of research becomes incessantly extensive, there is confusion regarding what type of blockchain or what part of the supply chain is addressed. This paper emphasizes some essential points to deal with much of the dubiety under consideration.
{"title":"Blockchain Technology between Nakamoto and Supply Chain Management: Insights from Academia and Practice","authors":"Rami Alkhudary","doi":"10.2139/ssrn.3660342","DOIUrl":"https://doi.org/10.2139/ssrn.3660342","url":null,"abstract":"Purpose: This paper aims to participate in answering the following research question: Will blockchain become the norm in supply chain management? Valuable insights from academia and industry about the use of blockchain technology in supply chain management are presented.<br><br>Design/methodology/approach: I review the literature that mainly examined the use of blockchain technology in supply chain management from 2008 to 2020. Additionally, I explore two blockchain projects: Nimble and Carrefour.<br><br>Findings: The mainstream research on the use of blockchain technology in supply chain management exhibits a positive trend. This study shows that relevant academic research increased by 174% in only 11 months (from February 2018 to January 2019). Countries and supply chains, to the same extent, are making bold moves toward using blockchain on a large scale. The European Commission, for example, funded a project to create a federated, multi-sided business ecosystem based on cloud services. The project developed a platform, named Nimble, to enable supply chain logistics to benefit from internet platforms and blockchain services. In another example, Carrefour has implemented blockchain in nine sectors (chicken, tomatoes, eggs, oranges, etc.) with a target of 300 sectors by 2022.<br><br>Originality/value: Although this area of research becomes incessantly extensive, there is confusion regarding what type of blockchain or what part of the supply chain is addressed. This paper emphasizes some essential points to deal with much of the dubiety under consideration.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"9 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83519573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work focuses on solving the data-driven contextual newsvendor problem with intertemporal dependence and non-stationarities. More specifically, we investigate learn the data-to-decision mapping for the newsvendor problem when observations of contexts and demands are available. The observations of both contexts and demands are generated sequentially in a fluctuate nature, thus exhibit an intertemporal dependence and even non-stationarities. However, most existing works that investigate the data-driven conditional Newsvendor problem adopt a common assumption that the data are independent and identically distributed (i.i.d.) to obtain performance guarantees such as generalization bounds. In this work, we develop performance guarantees in the form of out-of-sample generalization bounds for learning contextual newsvendor problem under comparatively more realistic assumptions including intertemporal dependence and moderate non-stationarities.
{"title":"Learning Newsvendor Problem with Intertemporal Dependence and Moderate Non-stationarities","authors":"Meng Qi, Z. Shen, Zeyu Zheng","doi":"10.2139/ssrn.3648615","DOIUrl":"https://doi.org/10.2139/ssrn.3648615","url":null,"abstract":"This work focuses on solving the data-driven contextual newsvendor problem with intertemporal dependence and non-stationarities. More specifically, we investigate learn the data-to-decision mapping for the newsvendor problem when observations of contexts and demands are available. The observations of both contexts and demands are generated sequentially in a fluctuate nature, thus exhibit an intertemporal dependence and even non-stationarities. However, most existing works that investigate the data-driven conditional Newsvendor problem adopt a common assumption that the data are independent and identically distributed (i.i.d.) to obtain performance guarantees such as generalization bounds. In this work, we develop performance guarantees in the form of out-of-sample generalization bounds for learning contextual newsvendor problem under comparatively more realistic assumptions including intertemporal dependence and moderate non-stationarities.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"30 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85138917","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rahul C. Basole, E. Bendoly, Aravind Chandrasekaran, K. Linderman
The unprecedented availability of data, along with the growing variety of software packages to visualize it, presents both opportunities and challenges for operations management (OM) research. OM researchers typically use data to describe conditions, predict phenomena, or make prescriptions depending on whether they are building, testing, or translating theories to practice. Visualization, when used appropriately, can complement, aid, and augment the researcher’s understanding in the different stages of research (theory building, testing, or translating and conveying results). On the other hand, if used incorrectly or without sufficient consideration, visualization can yield misleading and erroneous claims. This article formally examines the benefits of visualization as a complementary method enhancing each stage of a broader OM research strategy by examining frameworks and cases from extant research in different OM contexts. Our discussion offers guidance with regard to researchers’ use of visual data renderings, particularly toward avoiding misrepresentation, which can arise with the incorrect use of visualization. We close with a consideration of emerging trends and their implications for researchers and practitioners as well as recommendations for both authors and reviewers, regardless of domain, in evaluating the effectiveness of visuals at each stage of research.
{"title":"Visualization in Operations Management Research","authors":"Rahul C. Basole, E. Bendoly, Aravind Chandrasekaran, K. Linderman","doi":"10.2139/ssrn.3637040","DOIUrl":"https://doi.org/10.2139/ssrn.3637040","url":null,"abstract":"The unprecedented availability of data, along with the growing variety of software packages to visualize it, presents both opportunities and challenges for operations management (OM) research. OM researchers typically use data to describe conditions, predict phenomena, or make prescriptions depending on whether they are building, testing, or translating theories to practice. Visualization, when used appropriately, can complement, aid, and augment the researcher’s understanding in the different stages of research (theory building, testing, or translating and conveying results). On the other hand, if used incorrectly or without sufficient consideration, visualization can yield misleading and erroneous claims. This article formally examines the benefits of visualization as a complementary method enhancing each stage of a broader OM research strategy by examining frameworks and cases from extant research in different OM contexts. Our discussion offers guidance with regard to researchers’ use of visual data renderings, particularly toward avoiding misrepresentation, which can arise with the incorrect use of visualization. We close with a consideration of emerging trends and their implications for researchers and practitioners as well as recommendations for both authors and reviewers, regardless of domain, in evaluating the effectiveness of visuals at each stage of research.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"95 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84951620","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We compare three different approaches studied by past literature on data-driven inventory optimization--- Frequentist Parametric (FP), Bayesian Parametric (BP) and Nonparametric--- for the newsvendor problem. For the Parametric approaches, we allow for mis-specification of the demand model. We prove, under mild regularity conditions, (i) asymptotic bias and variance formulas of FP and BP are equivalent, (ii) mis-specified Parametric approaches yield asymptotically biased decisions, unlike the correctly-specified Parametric approaches and the Nonparametric approach, and (iii) asymptotic variance of the mis-specified Parametric approaches converges to zero at rate $1/n$, in contrast to the $1/n^2$ rate for the correctly-specified Parametric approaches and the Nonparametric approach, where $n$ is the number of demand samples. We then show, for nine pairs of assumed versus true demand distribution pairs, (iv) asymptotic bias and variance formulas approximate finite-sample counterparts very well, (v) correctly-specified Parametric approaches dominate the Nonparametric approach in the asymptotic mean-squared error (AMSE) of the decision and the cost, and (vi) surprisingly, it is possible for mis-specified Parametric approaches to dominate the Nonparametric approach in the AMSE of the decision and the cost. We compare the approaches on a dataset from a large fresh food chain, and discuss the nuances of choosing the ``best'' approach.
{"title":"Model Mis-Specification in Newsvendor Decisions: A Comparison of Frequentist Parametric, Bayesian Parametric and Nonparametric Approaches","authors":"Gah-Yi Ban, Zhenyu Gao, Fabian Taigel","doi":"10.2139/ssrn.3495733","DOIUrl":"https://doi.org/10.2139/ssrn.3495733","url":null,"abstract":"We compare three different approaches studied by past literature on data-driven inventory optimization--- Frequentist Parametric (FP), Bayesian Parametric (BP) and Nonparametric--- for the newsvendor problem. For the Parametric approaches, we allow for mis-specification of the demand model. We prove, under mild regularity conditions, (i) asymptotic bias and variance formulas of FP and BP are equivalent, (ii) mis-specified Parametric approaches yield asymptotically biased decisions, unlike the correctly-specified Parametric approaches and the Nonparametric approach, and (iii) asymptotic variance of the mis-specified Parametric approaches converges to zero at rate $1/n$, in contrast to the $1/n^2$ rate for the correctly-specified Parametric approaches and the Nonparametric approach, where $n$ is the number of demand samples. We then show, for nine pairs of assumed versus true demand distribution pairs, (iv) asymptotic bias and variance formulas approximate finite-sample counterparts very well, (v) correctly-specified Parametric approaches dominate the Nonparametric approach in the asymptotic mean-squared error (AMSE) of the decision and the cost, and (vi) surprisingly, it is possible for mis-specified Parametric approaches to dominate the Nonparametric approach in the AMSE of the decision and the cost. We compare the approaches on a dataset from a large fresh food chain, and discuss the nuances of choosing the ``best'' approach.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"14 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78305864","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Problem Definition: We investigate the implications of procedurally rational customers on service networks where customers visit multiple stations but can choose the order in which to visit the stations. Academic/Practical Relevance: Self-interested customers populate various service systems. While self-interested, these customers may not be fully rational. Customers' form of reasoning and its consequences for system performance affect the planning decisions of service providers. Methodology: We study procedurally rational customers---that is, customers make decisions based on anecdotal samples of system times experienced by customers who previously visited the system and followed each possible route. Using a fluid model, we fully characterize the evolution of customer routing decisions, with customers deciding in each period based on samples from the previous period. Results: We completely specify the set of equilibrium routing profiles, where the fraction of customers choosing each route becomes stationary. In contrast with existing models of procedural rationality, we find that procedurally rational customers sometimes behave differently from fully rational customers, but not always. Equilibria can emerge under procedural rationality that differ from fully rational equilibria, in which case, system performance suffers. We also study systems in which customers make routing decisions. We find analytically that procedurally rational customers are slower to internalize price increases than fully rational customers: accordingly, the firm's optimal revenue is higher (lower) with procedurally rational customers than fully rational ones if the waiting cost is high (low). Managerial Implications: In systems in which customers make only routing decisions, procedural rationality can lead to much longer waiting times. However, the firm can avoid increased waiting times if it can choose the service rates at its stations. Furthermore, if customers make joining decisions, then a service provider should price differently depending on which form of reasoning customers use, and the firm may prefer either procedurally rational customers or fully rational ones depending on the system parameters.
{"title":"Service Networks with Open Routing and Procedurally Rational Customers","authors":"Andrew E. Frazelle, Tingliang Huang, Y. Wei","doi":"10.2139/ssrn.3414468","DOIUrl":"https://doi.org/10.2139/ssrn.3414468","url":null,"abstract":"Problem Definition: We investigate the implications of procedurally rational customers on service networks where customers visit multiple stations but can choose the order in which to visit the stations. \u0000 \u0000Academic/Practical Relevance: Self-interested customers populate various service systems. While self-interested, these customers may not be fully rational. Customers' form of reasoning and its consequences for system performance affect the planning decisions of service providers. \u0000Methodology: We study procedurally rational customers---that is, customers make decisions based on anecdotal samples of system times experienced by customers who previously visited the system and followed each possible route. Using a fluid model, we fully characterize the evolution of customer routing decisions, with customers deciding in each period based on samples from the previous period. \u0000 \u0000Results: We completely specify the set of equilibrium routing profiles, where the fraction of customers choosing each route becomes stationary. In contrast with existing models of procedural rationality, we find that procedurally rational customers sometimes behave differently from fully rational customers, but not always. Equilibria can emerge under procedural rationality that differ from fully rational equilibria, in which case, system performance suffers. We also study systems in which customers make routing decisions. We find analytically that procedurally rational customers are slower to internalize price increases than fully rational customers: accordingly, the firm's optimal revenue is higher (lower) with procedurally rational customers than fully rational ones if the waiting cost is high (low). \u0000 \u0000Managerial Implications: In systems in which customers make only routing decisions, procedural rationality can lead to much longer waiting times. However, the firm can avoid increased waiting times if it can choose the service rates at its stations. Furthermore, if customers make joining decisions, then a service provider should price differently depending on which form of reasoning customers use, and the firm may prefer either procedurally rational customers or fully rational ones depending on the system parameters.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"40 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81013837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As digitization enables service platforms to track users’ footprints and control information design, important questions arise about how digital service platforms should disseminate customer-related information to service providers to shape user behavior in the desired direction. We contribute to this understanding by examining the impact of a novel information design — providing customer-related information to facilitate interaction between service providers and customers at the beginning of a service encounter — on service capacity and quality. We evaluate this strategy via a field experiment on a live streaming platform that connects individual broadcasters and viewers. When viewers entered shows, we provided information about viewers to broadcasters that were randomly assigned to the treatment condition (but not to control broadcasters). Our analysis involving 49,999 broadcasters confirms that our intervention increased the interaction between broadcasters and viewers by 8.66% during the first minute after viewers entered a show. Furthermore, treatment broadcasters expanded their service capacity by 8.74% via increasing both the frequency (by 3.96%) and length (by 7.52%) of shows. This is not because treatment broadcasters could or believed they could earn more per minute of service but rather because they felt more connected with viewers and enjoyed performing more. Moreover, our intervention increased experienced service quality (as measured by viewers’ watch time) by 5.85%, without affecting willingness to pay. The increase is not simply driven by treatment broadcasters’ expanded service capacity. Instead, viewers were more likely to ``follow" treatment broadcasters and watch their future shows, suggesting that our intervention made viewers feel more connected with broadcasters. Our low-cost information-based intervention has important implications for digital service platforms that do not have control over service providers’ work schedules and service quality.
{"title":"Information Design to Facilitate Social Interactions on Service Platforms: Evidence from a Large Field Experiment","authors":"Hengchen Dai, Dennis J. Zhang, Zhiwei Xu","doi":"10.2139/ssrn.3528619","DOIUrl":"https://doi.org/10.2139/ssrn.3528619","url":null,"abstract":"As digitization enables service platforms to track users’ footprints and control information design, important questions arise about how digital service platforms should disseminate customer-related information to service providers to shape user behavior in the desired direction. We contribute to this understanding by examining the impact of a novel information design — providing customer-related information to facilitate interaction between service providers and customers at the beginning of a service encounter — on service capacity and quality. We evaluate this strategy via a field experiment on a live streaming platform that connects individual broadcasters and viewers. When viewers entered shows, we provided information about viewers to broadcasters that were randomly assigned to the treatment condition (but not to control broadcasters). Our analysis involving 49,999 broadcasters confirms that our intervention increased the interaction between broadcasters and viewers by 8.66% during the first minute after viewers entered a show. Furthermore, treatment broadcasters expanded their service capacity by 8.74% via increasing both the frequency (by 3.96%) and length (by 7.52%) of shows. This is not because treatment broadcasters could or believed they could earn more per minute of service but rather because they felt more connected with viewers and enjoyed performing more. Moreover, our intervention increased experienced service quality (as measured by viewers’ watch time) by 5.85%, without affecting willingness to pay. The increase is not simply driven by treatment broadcasters’ expanded service capacity. Instead, viewers were more likely to ``follow\" treatment broadcasters and watch their future shows, suggesting that our intervention made viewers feel more connected with broadcasters. Our low-cost information-based intervention has important implications for digital service platforms that do not have control over service providers’ work schedules and service quality.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"44 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89391898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Classical facility location problems do not incorporate the possibility of disruptions among facilities and usually result in solutions that do not perform well under disruptions. Existent literature on disreputable facility locations focuses on independent or extreme dependence scenarios where the probability structure is simplistic but has the advantage of allowing for efficient optimization. We propose the use of partially subordinated Markov Chains to model the probability of the dependent risk of disruptions. This parsimonious approach o ers a realistic model for disruptions. We also propose algorithms to calibrate a partially subordinated Markov Chain model and to optimize for the facility locations under dependent disruptions. We calibrate our model on two different cases with four different disruption data sets each and solve for the optimal facility location choice. We show that the resulting solutions significantly outperform those yielded from stronger assumptions like independence or extreme dependence.
{"title":"Facility Location with Joint Disruptions","authors":"Vishwakant Malladi, K. Muthuraman","doi":"10.2139/ssrn.3515230","DOIUrl":"https://doi.org/10.2139/ssrn.3515230","url":null,"abstract":"Classical facility location problems do not incorporate the possibility of disruptions among facilities and usually result in solutions that do not perform well under disruptions. Existent literature on disreputable facility locations focuses on independent or extreme dependence scenarios where the probability structure is simplistic but has the advantage of allowing for efficient optimization. We propose the use of partially subordinated Markov Chains to model the probability of the dependent risk of disruptions. This parsimonious approach o ers a realistic model for disruptions. We also propose algorithms to calibrate a partially subordinated Markov Chain model and to optimize for the facility locations under dependent disruptions. We calibrate our model on two different cases with four different disruption data sets each and solve for the optimal facility location choice. We show that the resulting solutions significantly outperform those yielded from stronger assumptions like independence or extreme dependence.","PeriodicalId":49886,"journal":{"name":"Manufacturing Engineering","volume":"169 1","pages":""},"PeriodicalIF":0.3,"publicationDate":"2020-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90430542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：