Motivated by the observation that overexposure to unwanted marketing activities can lead to customer dissatisfaction, we consider a setting where a platform offers a sequence of messages to its users and is penalized when users abandon the platform due to marketing fatigue. We propose a novel sequential choice model to capture multiple interactions taking place between the platform and its users: upon receiving a message, a user decides on whether to accept or reject the message. If she chooses to reject, she would then decide to either receive the next message in the sequence or abandon the platform. Based on user feedback, the platform dynamically learns users' abandonment distribution and the relevance of the recommended content. With a goal to maximize the cumulative payoff over a horizon of length T, the platform dynamically adjusts the sequence of messages and the order in which the messages are shown to a user. We refer to this online learning task as the sequential choice bandit (SC-Bandit) problem. For the offline combinatorial optimization problem, we show a polynomial-time algorithm. For the online problem, we consider two variants, depending on whether contexts are included, and propose algorithms that balance exploration and exploitation. Lastly, we evaluate the performance of our algorithms with both synthetic and real-world datasets.
{"title":"Sequential Choice Bandits: Learning with Marketing Fatigue","authors":"Junyu Cao, Wei Sun, Z. Shen","doi":"10.2139/ssrn.3355211","DOIUrl":"https://doi.org/10.2139/ssrn.3355211","url":null,"abstract":"Motivated by the observation that overexposure to unwanted marketing activities can lead to customer dissatisfaction, we consider a setting where a platform offers a sequence of messages to its users and is penalized when users abandon the platform due to marketing fatigue. We propose a novel sequential choice model to capture multiple interactions taking place between the platform and its users: upon receiving a message, a user decides on whether to accept or reject the message. If she chooses to reject, she would then decide to either receive the next message in the sequence or abandon the platform. Based on user feedback, the platform dynamically learns users' abandonment distribution and the relevance of the recommended content. With a goal to maximize the cumulative payoff over a horizon of length T, the platform dynamically adjusts the sequence of messages and the order in which the messages are shown to a user. We refer to this online learning task as the sequential choice bandit (SC-Bandit) problem. For the offline combinatorial optimization problem, we show a polynomial-time algorithm. For the online problem, we consider two variants, depending on whether contexts are included, and propose algorithms that balance exploration and exploitation. Lastly, we evaluate the performance of our algorithms with both synthetic and real-world datasets.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130473778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Designing Fair and Efficient Matching Service Systems
设计公平高效的配对服务体系
{"title":"On the Optimal Design of a Bipartite Matching Queueing System","authors":"Philipp Afèche, René Caldentey, Varun Gupta","doi":"10.2139/ssrn.3345302","DOIUrl":"https://doi.org/10.2139/ssrn.3345302","url":null,"abstract":"Designing Fair and Efficient Matching Service Systems","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131937074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Jansen, F. Erhun, Nektarios Oraiopoulos, D. Ralph
Responding to an information technology (IT) system failure often requires a collaborative approach in which both the client and the vendor need to invest in response capacity. By investing more in response capacity, the client might make the vendor's response capacity more effective in the system restoration stage. Yet, in doing so, the client also encourages free-riding by the vendor. To understand how a client should balance the need to support the vendor while setting the right incentives for the vendor to invest, we develop a model that combines the key characteristics of value co-creation (i.e., complementarity between the firms' investments in response capacity) with standard maintenance contract practices (i.e., penalty-based contracts that penalize the vendor for system downtime). We study the value of observability by characterizing the difference in the client's expected utility between when her investment is observable and non-observable by the vendor in collaborative environments. Since exposure to increased financial risks is a critical issue for the vendors with performance-based contracts, we consider the impact of risk attitudes of the firms (i.e., vendor risk aversion (VRA) and client risk aversion (CRA)) on the investments in the collaborative response process. We show that the value of observability is decreasing in VRA but increasing in CRA. Secondly, we find that the effect of risk aversion on the average system downtime is diametrically opposite depending on whether or not the client's investment is observable. Finally, the effectiveness of the performance-based contracts decreases with VRA but is more robust to CRA.
{"title":"Enabling Collaborative Response to IT Service Disruptions Under Risk Aversion","authors":"M. Jansen, F. Erhun, Nektarios Oraiopoulos, D. Ralph","doi":"10.2139/ssrn.3332323","DOIUrl":"https://doi.org/10.2139/ssrn.3332323","url":null,"abstract":"Responding to an information technology (IT) system failure often requires a collaborative approach in which both the client and the vendor need to invest in response capacity. By investing more in response capacity, the client might make the vendor's response capacity more effective in the system restoration stage. Yet, in doing so, the client also encourages free-riding by the vendor. To understand how a client should balance the need to support the vendor while setting the right incentives for the vendor to invest, we develop a model that combines the key characteristics of value co-creation (i.e., complementarity between the firms' investments in response capacity) with standard maintenance contract practices (i.e., penalty-based contracts that penalize the vendor for system downtime). \u0000 \u0000We study the value of observability by characterizing the difference in the client's expected utility between when her investment is observable and non-observable by the vendor in collaborative environments. Since exposure to increased financial risks is a critical issue for the vendors with performance-based contracts, we consider the impact of risk attitudes of the firms (i.e., vendor risk aversion (VRA) and client risk aversion (CRA)) on the investments in the collaborative response process. We show that the value of observability is decreasing in VRA but increasing in CRA. Secondly, we find that the effect of risk aversion on the average system downtime is diametrically opposite depending on whether or not the client's investment is observable. Finally, the effectiveness of the performance-based contracts decreases with VRA but is more robust to CRA.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125439416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
There are in general three types of fees used by platforms: the fixed membership fee, transaction fee and proportional fee. In real-life, all three fees are used by different platforms. Often, the fee charged by the market maker is a combination of the fixed, transaction and proportional fees. We try to offer a rationale for the usage of these fee structures. Compared with the transaction fee, the profit for the platform is higher under the proportional fee. If the consumption value is high enough, then the profit under the membership fee is the lowest. If the value is low enough, then the profit under the membership fee is the highest. If the value is in the middle, then the profit under the membership fee is in the middle as well. However, when we allow free entry of merchants, results are different. The profit of the card network could be lower under the proportional fee than the transaction fee with free entry of merchants. Equilibrium under the proportional fee and the membership fee is the same. Consumer surplus is lower under the transaction fee than the proportional fee and the membership fee. But social welfare could be higher under the transaction fee. We also consider regulation of the fee structure and the optimal fee.
{"title":"The Proportional Fee, Transaction Fee and Membership Fee","authors":"Linfeng Chen","doi":"10.2139/ssrn.3320939","DOIUrl":"https://doi.org/10.2139/ssrn.3320939","url":null,"abstract":"There are in general three types of fees used by platforms: the fixed membership fee, transaction fee and proportional fee. In real-life, all three fees are used by different platforms. Often, the fee charged by the market maker is a combination of the fixed, transaction and proportional fees. We try to offer a rationale for the usage of these fee structures. Compared with the transaction fee, the profit for the platform is higher under the proportional fee. If the consumption value is high enough, then the profit under the membership fee is the lowest. If the value is low enough, then the profit under the membership fee is the highest. If the value is in the middle, then the profit under the membership fee is in the middle as well. However, when we allow free entry of merchants, results are different. The profit of the card network could be lower under the proportional fee than the transaction fee with free entry of merchants. Equilibrium under the proportional fee and the membership fee is the same. Consumer surplus is lower under the transaction fee than the proportional fee and the membership fee. But social welfare could be higher under the transaction fee. We also consider regulation of the fee structure and the optimal fee.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121958429","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Standard economic theory suggests that monopolies result in outputs lower and prices higher than socially desirable. In service systems, customers are often reluctant to join overly crowded systems because their service valuation decreases with system congestion. Thus a high service price is associated with better service through low congestion levels, that is, low system output. But can a monopolist profit more by providing lots of customers with poor service for a very low price? In our work, we introduce a unified approach, relying on the concept of observable queues, for studying the phenomena of monopoly overpricing in service systems. We explain why, in most observable queue models, the monopolist tends to underexploit capacity by overcharging its service. Yet we discuss cases in which the monopolist may prefer to attract demand by charging less than the socially optimal price.
{"title":"Social and Monopoly Optimization in Observable Queues","authors":"Refael Hassin, Ran I. Snitkovsky","doi":"10.2139/ssrn.3290251","DOIUrl":"https://doi.org/10.2139/ssrn.3290251","url":null,"abstract":"Standard economic theory suggests that monopolies result in outputs lower and prices higher than socially desirable. In service systems, customers are often reluctant to join overly crowded systems because their service valuation decreases with system congestion. Thus a high service price is associated with better service through low congestion levels, that is, low system output. But can a monopolist profit more by providing lots of customers with poor service for a very low price? In our work, we introduce a unified approach, relying on the concept of observable queues, for studying the phenomena of monopoly overpricing in service systems. We explain why, in most observable queue models, the monopolist tends to underexploit capacity by overcharging its service. Yet we discuss cases in which the monopolist may prefer to attract demand by charging less than the socially optimal price.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"184 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132967083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kjartan Kloster Osmundsen, T. S. Kleppe, R. Liesenfeld
The joint posterior of latent variables and parameters in Bayesian hierarchical models often has a strong nonlinear dependence structure, thus making it a challenging target for standard Markov-chain Monte-Carlo methods. Pseudo-marginal methods aim at effectively exploring such target distributions, by marginalizing the latent variables using Monte-Carlo integration and directly targeting the marginal posterior of the parameters. We follow this approach and propose a generic pseudo-marginal algorithm for efficiently simulating from the posterior of the parameters. It combines efficient importance sampling, for accurately marginalizing the latent variables, with the recently developed pseudo-marginal Hamiltonian Monte Carlo approach. We illustrate our algorithm in applications to dynamic state space models, where it shows a very high simulation efficiency even in challenging scenarios with complex dependence structures.
{"title":"Pseudo-Marginal Hamiltonian Monte Carlo with Efficient Importance Sampling","authors":"Kjartan Kloster Osmundsen, T. S. Kleppe, R. Liesenfeld","doi":"10.2139/ssrn.3304077","DOIUrl":"https://doi.org/10.2139/ssrn.3304077","url":null,"abstract":"The joint posterior of latent variables and parameters in Bayesian hierarchical models often has a strong nonlinear dependence structure, thus making it a challenging target for standard Markov-chain Monte-Carlo methods. Pseudo-marginal methods aim at effectively exploring such target distributions, by marginalizing the latent variables using Monte-Carlo integration and directly targeting the marginal posterior of the parameters. We follow this approach and propose a generic pseudo-marginal algorithm for efficiently simulating from the posterior of the parameters. It combines efficient importance sampling, for accurately marginalizing the latent variables, with the recently developed pseudo-marginal Hamiltonian Monte Carlo approach. We illustrate our algorithm in applications to dynamic state space models, where it shows a very high simulation efficiency even in challenging scenarios with complex dependence structures.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117047836","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Adam N. Elmachtoub, Vishal Gupta, Michael L. Hamilton
Increased availability of high-quality customer information has fueled interest in personalized pricing strategies, that is, strategies that predict an individual customer’s valuation for a product and then offer a price tailored to that customer. Although the appeal of personalized pricing is clear, it may also incur large costs in the forms of market research, investment in information technology and analytics expertise, and branding risks. In light of these trade-offs, our work studies the value of personalized pricing strategies over a simple single-price strategy. We first provide closed-form lower and upper bounds on the ratio between the profits of an idealized personalized pricing strategy (first-degree price discrimination) and a single-price strategy. Our bounds depend on simple statistics of the valuation distribution and shed light on the types of markets for which personalized pricing has little or significant potential value. Second, we consider a feature-based pricing model where customer valuations can be estimated from observed features. We show how to transform our aforementioned bounds into lower and upper bounds on the value of feature-based pricing over single pricing depending on the degree to which the features are informative for the valuation. Finally, we demonstrate how to obtain sharper bounds by incorporating additional information about the valuation distribution (moments or shape constraints) by solving tractable linear optimization problems. This paper was accepted by David Simchi-Levi, revenue management and market analytics.
{"title":"The Value of Personalized Pricing","authors":"Adam N. Elmachtoub, Vishal Gupta, Michael L. Hamilton","doi":"10.2139/ssrn.3127719","DOIUrl":"https://doi.org/10.2139/ssrn.3127719","url":null,"abstract":"Increased availability of high-quality customer information has fueled interest in personalized pricing strategies, that is, strategies that predict an individual customer’s valuation for a product and then offer a price tailored to that customer. Although the appeal of personalized pricing is clear, it may also incur large costs in the forms of market research, investment in information technology and analytics expertise, and branding risks. In light of these trade-offs, our work studies the value of personalized pricing strategies over a simple single-price strategy. We first provide closed-form lower and upper bounds on the ratio between the profits of an idealized personalized pricing strategy (first-degree price discrimination) and a single-price strategy. Our bounds depend on simple statistics of the valuation distribution and shed light on the types of markets for which personalized pricing has little or significant potential value. Second, we consider a feature-based pricing model where customer valuations can be estimated from observed features. We show how to transform our aforementioned bounds into lower and upper bounds on the value of feature-based pricing over single pricing depending on the degree to which the features are informative for the valuation. Finally, we demonstrate how to obtain sharper bounds by incorporating additional information about the valuation distribution (moments or shape constraints) by solving tractable linear optimization problems. This paper was accepted by David Simchi-Levi, revenue management and market analytics.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"242 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114634074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study the assortment optimization problem in an online setting where a retailer uses multiple distribution centers to fulfill customer orders. Due to space, handling or other constraints, each distribution center can carry up to a pre-specified number of products. It is assumed that each distribution center is primarily responsible for a geographical region whose customers' choice is governed by a separate multinomial logit model. A distribution center can satisfy the demand from other regions, but this incurs an additional shipping cost for the retailer. The problem for the retailer is to determine which products to carry in each of its distribution centers and which products to offer for sale in each region so as to maximize its expected profit (revenue minus the shipping costs). We first show that the problem is NP-complete. We develop a conic quadratic mixed integer programming formulation and suggest a family of valid inequalities to strengthen this formulation. Numerical experiments show that our conic approach, combined with valid inequalities over-perform the mixed integer linear programming formulation and enables us to solve moderately sized instances optimally. We also study the effect of various factors such as the strength of the outside option, capacity constraint and shipping cost on company's profitability and assortment selection. Finally, we study the effect of not allowing cross-shipments or not considering them in assortment decisions and show that these may lead to substantial losses for an online retailer.
{"title":"Multi-Location Assortment Optimization Under Capacity Constraints","authors":"Başak Bebitoğlu, Alper Şen, Philip M. Kaminsky","doi":"10.2139/ssrn.3249175","DOIUrl":"https://doi.org/10.2139/ssrn.3249175","url":null,"abstract":"We study the assortment optimization problem in an online setting where a retailer uses multiple distribution centers to fulfill customer orders. Due to space, handling or other constraints, each distribution center can carry up to a pre-specified number of products. It is assumed that each distribution center is primarily responsible for a geographical region whose customers' choice is governed by a separate multinomial logit model. A distribution center can satisfy the demand from other regions, but this incurs an additional shipping cost for the retailer. The problem for the retailer is to determine which products to carry in each of its distribution centers and which products to offer for sale in each region so as to maximize its expected profit (revenue minus the shipping costs). We first show that the problem is NP-complete. We develop a conic quadratic mixed integer programming formulation and suggest a family of valid inequalities to strengthen this formulation. Numerical experiments show that our conic approach, combined with valid inequalities over-perform the mixed integer linear programming formulation and enables us to solve moderately sized instances optimally. We also study the effect of various factors such as the strength of the outside option, capacity constraint and shipping cost on company's profitability and assortment selection. Finally, we study the effect of not allowing cross-shipments or not considering them in assortment decisions and show that these may lead to substantial losses for an online retailer.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132828641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Saed Alizamir, Ningyuan Chen, Sang‐Hyun Kim, V. Manshadi
We analyze a firm’s optimal pricing of a new service when consumers interact in a network and impose positive externality on one another. The firm initially provides its service for free, leveraging network externality to promote rapid service consumption growth. The firm raises the price and starts earning revenue only when a sufficient level of consumer interactions is established. Incorporating the local network effects in a non-stationary dynamic model, we study the impact of network structure on the firm’s revenue and optimal pricing decision. We find that the firm delays the timing of service monetization when it faces a more strongly connected network, despite the fact that such a network allows the firm to monetize the service sooner by resulting in faster consumption growth. We also find that the firm benefits from network imbalance, i.e., the firm prefers a network of consumers with varying degrees of connections to that with similar degrees of connections. We also study the value of knowing the network structure and discuss how such knowledge impacts the firm’s profitability. Our analyses rely on the techniques from algebraic graph theory which enable us to solve the firm’s high-dimensional dynamic pricing problem by relating it to the network’s spectral characteristics.
{"title":"Impact of Network Structure on New Service Pricing","authors":"Saed Alizamir, Ningyuan Chen, Sang‐Hyun Kim, V. Manshadi","doi":"10.2139/ssrn.3236225","DOIUrl":"https://doi.org/10.2139/ssrn.3236225","url":null,"abstract":"We analyze a firm’s optimal pricing of a new service when consumers interact in a network and impose positive externality on one another. The firm initially provides its service for free, leveraging network externality to promote rapid service consumption growth. The firm raises the price and starts earning revenue only when a sufficient level of consumer interactions is established. Incorporating the local network effects in a non-stationary dynamic model, we study the impact of network structure on the firm’s revenue and optimal pricing decision. We find that the firm delays the timing of service monetization when it faces a more strongly connected network, despite the fact that such a network allows the firm to monetize the service sooner by resulting in faster consumption growth. We also find that the firm benefits from network imbalance, i.e., the firm prefers a network of consumers with varying degrees of connections to that with similar degrees of connections. We also study the value of knowing the network structure and discuss how such knowledge impacts the firm’s profitability. Our analyses rely on the techniques from algebraic graph theory which enable us to solve the firm’s high-dimensional dynamic pricing problem by relating it to the network’s spectral characteristics.","PeriodicalId":275253,"journal":{"name":"Operations Research eJournal","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128379434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","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：