Abstract
Pricing and Resource Allocation Made Simple for Service Systems
Large-scale service systems—from drone delivery to cloud computing—face the dual challenge of balancing customer delays with revenue maximization. In “Near-Optimal Pricing and Resource Allocation in a Large-Scale Service System,” Wu, Liu, and Sun propose a dual-based pricing and resource allocation policy that is both simple and theoretically powerful. This greedy, one-step heuristic delivers performance guarantees matching the theoretical lower bound.
Beyond the stylized model that illustrates its core idea, the study shows the value of dynamic pricing through an insensitivity result: any work-conserving rule can stabilize the system. The policy also proves robust under realistic conditions, including heterogeneous server pools and nonexponential service environments.
Perhaps most striking, the authors uncover a “compensation effect”: near-optimal pricing curves need not rise monotonically with congestion. Instead, they may offset customers’ delay disutility to sustain revenue. These insights offer a practical, theory-backed framework for modern service operations.
We study dynamic pricing and resource allocation in large-scale service systems where multiple service units serve customers who are both price and delay sensitive. Customers are segmented into classes characterized by class-specific service rates and demand functions shaped by posted prices and estimated delays. To jointly optimize revenue and delay performance, we propose a family of state-dependent greedy heuristics that (i) assign dedicated service capacities to each customer class, and (ii) dynamically set prices by solving a tractable one-step optimization problem. Despite their simplicity, these heuristics achieve a relative optimality gap of [Formula: see text] in large-market regimes where demand scales with the number of servers n. We further establish that replacing the dedicated-capacity rule with any work-conserving allocation preserves the same order of optimality. Numerical experiments confirm the efficacy and robustness of our approach and offer additional insights, including the counterintuitive finding that congestion-based pricing can be nonmonotonic in system congestion when customers observe queue-based wait-time estimates.
Funding: This research was supported by the National Natural Science Foundation of China [Grants 72371161 and 71972133].
Supplemental Material: All supplemental materials, including the code, data, and files required to reproduce the results, are available at https://doi.org/10.1287/opre.2024.1073 .