About

Achal Bassamboo is the Charles E. Morrison Professor at the Kellogg School of Management, Northwestern University, and serves as the co-director of the MMM program, a dual degree between Kellogg and the Segal Design at McCormick School. He joined the Kellogg faculty in 2005 after completing his Ph.D. in Operations, Information, and Technology at Stanford Graduate School of Business. His research interests include service systems, revenue management, and information sharing, with current work focusing on designing flexible service systems, capacity planning under parameter uncertainty, and studying the credibility of information provided by service providers or retailers. Professor Bassamboo has published articles in leading journals such as Management Science, Manufacturing and Service Operations Management, and Operations Research. He has received recognition for his research, including the 2016 Young Scholar Award from the Manufacturing and Service Operations Management Society, and has served on editorial boards for prominent journals including Management Science, POMS, and Naval Research Logistics. He teaches courses on operations management, supply chain logistics, decision models, and statistics, and his academic background includes a Ph.D. from Stanford University, a master's in Statistics from Stanford, and a B.Tech. in Mechanical Engineering from the Indian Institute of Technology.

Research topics

• Computer Science
• Marketing
• Business
• Economics
• Microeconomics
• Industrial organization
• Telecommunications
• Operations management
• Operations research
• Computer network
• Mathematics
• Econometrics
• Psychology
• Labour economics

Selected publications

• Generating DDPM-based Samples from Tilted Distributions

  arXiv (Cornell University) · 2026-04-03

  preprintOpen access

  Given $n$ independent samples from a $d$-dimensional probability distribution, our aim is to generate diffusion-based samples from a distribution obtained by tilting the original, where the degree of tilt is parametrized by $θ\in \mathbb{R}^d$. We define a plug-in estimator and show that it is minimax-optimal. We develop Wasserstein bounds between the distribution of the plug-in estimator and the true distribution as a function of $n$ and $θ$, illustrating regimes where the output and the desired true distribution are close. Further, under some assumptions, we prove the TV-accuracy of running Diffusion on these tilted samples. Our theoretical results are supported by extensive simulations. Applications of our work include finance, weather and climate modelling, and many other domains, where the aim may be to generate samples from a tilted distribution that satisfies practically motivated moment constraints.

  PublisherDOI

• When More Experience Is Not More Helpful: Evidence from Positive and Negative Reviews on Steam

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access
  PublisherDOI

• Generating DDPM-based Samples from Tilted Distributions

  ArXiv.org · 2026-04-03

  articleOpen access

  Given $n$ independent samples from a $d$-dimensional probability distribution, our aim is to generate diffusion-based samples from a distribution obtained by tilting the original, where the degree of tilt is parametrized by $θ\in \mathbb{R}^d$. We define a plug-in estimator and show that it is minimax-optimal. We develop Wasserstein bounds between the distribution of the plug-in estimator and the true distribution as a function of $n$ and $θ$, illustrating regimes where the output and the desired true distribution are close. Further, under some assumptions, we prove the TV-accuracy of running Diffusion on these tilted samples. Our theoretical results are supported by extensive simulations. Applications of our work include finance, weather and climate modelling, and many other domains, where the aim may be to generate samples from a tilted distribution that satisfies practically motivated moment constraints.

  PublisherOA PDF

• Fundamental limits for weighted empirical approximations of tilted distributions

  arXiv (Cornell University) · 2025-12-30

  preprintOpen access

  Consider the task of generating samples from a tilted distribution of a random vector whose underlying distribution is unknown, but samples from it are available. This finds applications in fields such as finance and climate science, and in rare event simulation. In this article, we discuss the asymptotic efficiency of a self-normalized importance sampler of the tilted distribution. We provide a sharp characterization of its accuracy, given the number of samples and the degree of tilt. Our findings reveal a surprising dichotomy: while the number of samples needed to accurately tilt a bounded random vector increases polynomially in the tilt amount, it increases at a super polynomial rate for unbounded distributions.

  PublisherDOI

• Fundamental limits for weighted empirical approximations of tilted distributions

  ArXiv.org · 2025-12-30

  articleOpen access

  Consider the task of generating samples from a tilted distribution of a random vector whose underlying distribution is unknown, but samples from it are available. This finds applications in fields such as finance and climate science, and in rare event simulation. In this article, we discuss the asymptotic efficiency of a self-normalized importance sampler of the tilted distribution. We provide a sharp characterization of its accuracy, given the number of samples and the degree of tilt. Our findings reveal a surprising dichotomy: while the number of samples needed to accurately tilt a bounded random vector increases polynomially in the tilt amount, it increases at a super polynomial rate for unbounded distributions.

  PublisherOA PDF

• The Queue Behind the Curtain: Information Disclosure in Omnichannel Services

  Naval Research Logistics (NRL) · 2025-08-16 · 1 citations

  article

  ABSTRACT With evolving mobile technologies, an increasing number of service providers are running multiple channels by offering apps to serve customers. In this article, we evaluate whether providing an online ordering option through an app is necessarily beneficial for the firm and customers. In particular, we study the implications of queue‐length information disclosure on the channel choice strategy of wait‐sensitive and quality‐sensitive app users. We adopt a game‐theoretic, discrete‐time queuing framework to model customers' strategic channel choice behavior in omnichannel systems. When provided with queue‐length information in an omnichannel system, app users follow a dual‐threshold policy where they order online for moderate queue lengths and choose the offline option when the queue length is either too short or too long. If non‐app users are relatively more wait‐sensitive, the overall throughput might be lower in the omnichannel system compared to the single‐channel benchmark. On the other hand, if app users are relatively more wait‐sensitive, then the omnichannel system increases throughput. From the customer's perspective, whether or not providing online ordering benefits her depends on the relative market size and the relative wait‐sensitivities of the non‐app user and app user segments. While non‐app users are consistently worse off in omnichannel systems, app users certainly benefit from omnichannel systems when queue‐length information is disclosed. It is indeed possible that both segments might be worse off in an omnichannel system. The firm may use information disclosure in an omnichannel system as an operational lever to increase throughput when app users are either highly quality‐sensitive or highly wait‐sensitive, or when the system is highly congested. Evaluating the overall performance of an omnichannel firm requires a careful calibration of customer primitives and consideration of the relative proportions of non‐app users and app users in the system.

  PublisherDOI

• Asymptotic optimality theory of confidence intervals of the mean

  ArXiv.org · 2025-01-31

  preprintOpen access

  We address the classical problem of constructing confidence intervals (CIs) for the mean of a distribution, given \(N\) i.i.d. samples, such that the CI contains the true mean with probability at least \(1 - δ\), where \(δ\in (0,1)\). We characterize three distinct learning regimes based on the minimum achievable limiting width of any CI as the sample size \(N_δ \to \infty\) and \(δ\to 0\). In the first regime, where \(N_δ\) grows slower than \(\log(1/δ)\), the limiting width of any CI equals the width of the distribution's support, precluding meaningful inference. In the second regime, where \(N_δ\) scales as \(\log(1/δ)\), we precisely characterize the minimum limiting width, which depends on the scaling constant. In the third regime, where \(N_δ\) grows faster than \(\log(1/δ)\), complete learning is achievable, and the limiting width of the CI collapses to zero, converging to the true mean. We demonstrate that CIs derived from concentration inequalities based on Kullback--Leibler (KL) divergences achieve asymptotically optimal performance, attaining the minimum limiting width in both sufficient and complete learning regimes for distributions in two families: single-parameter exponential and bounded support. Additionally, these results extend to one-sided CIs, with the width notion adjusted appropriately. Finally, we generalize our findings to settings with random per-sample costs, motivated by practical applications such as stochastic simulators and cloud service selection. Instead of a fixed sample size, we consider a cost budget \(C_δ\), identifying analogous learning regimes and characterizing the optimal CI construction policy.

  PublisherOA PDFDOI

• Adapting to Gig Economy Dynamics: Network and Staffing Strategies for Long-Haul Trucking Platforms

  SSRN Electronic Journal · 2024-01-01

  preprintOpen accessSenior author
  PublisherDOI

• Is Full Price the Full Story When Consumers Have Time and Budget Constraints?

  Manufacturing & Service Operations Management · 2023 · 2 citations

  Senior authorCorresponding
  • Microeconomics
  • Economics
  • Business

  Problem definition: A canonical model in service management assumes that consumers base the purchase of a service on its full price, that is, a linear combination of the monetary price and the expected time commitment. Although analytically convenient, when this assumption holds is an unexplored question. Methodology/results: We present a model of consumers allocating their time and money between working, overhead activities that do not provide utility, one continuous leisure activity, and one discrete service. Both continuous leisure activity and discrete service increase utility. Consumers can allocate any nonnegative amount of time or money to the leisure activity. Consumption of the discrete service requires a specific amount of time and money. We examine when the decision to purchase the discrete service depends only on its full price. We show that the full-price assumption does hold in specific cases. To be precise, it depends on how consumers are paid. If consumers completely control the amount of time that they work and earn a constant wage, they base their purchase decision on the full price. If, however, they must work a fixed shift length, then the assumption fails, and the full price is not sufficient to determine the consumer’s action. This leads to systematic differences in sellers’ strategies when they serve consumers with different compensation structures. If the consumers must work longer than would be optimal if they controlled their schedule and earned the same hourly wage, that is, the consumers are overemployed shift workers, then a seller restricts sales (relative to selling to consumers who control their work hours), and the system is less congested. The reverse holds if the consumers would prefer to work longer at the offered wage; that is, the consumers are underemployed shift workers. Managerial implications: We show that sellers who fail to take prevailing compensation structures of the community they serve into consideration experience significant revenue loss. In some cases, we see losses in consumer surplus and social welfare as well. Supplemental Material: The e-companion is available at https://doi.org/10.1287/msom.2022.0357 .

  PublisherDOI

• Cents of Urgency: How Opening a Co-located Urgent Care Center Affects Emergency Department Arrivals

  SSRN Electronic Journal · 2023-01-01

  articleOpen access
  PublisherDOI

Frequent coauthors

• Gad Allon

  University of Pennsylvania

  30 shared

• Ramandeep S. Randhawa

  University of Southern California

  20 shared

• Assaf Zeevi

  Columbia University

  17 shared

• Chenguang Wu

  Hong Kong University of Science and Technology

  13 shared

• Sandeep Juneja

  TB Alliance

  11 shared

• Sunil Chopra

  Northwestern University

  10 shared

• Mark S. Daskin

  7 shared

• Ioannis Stamatopoulos

  The University of Texas at Austin

  7 shared

Awards & honors

• 2016 "Young Scholar Award" by the Manufacturing and Service…