Job Market Paper
- "Targeting with Operational Constraints"Job Market Paper | Draft available upon request
Abstract
Most service firms operate under capacity constraints that limit how effectively they can respond to demand fluctuations. When targeted marketing interventions stimulate demand, they can unintentionally overload operational capacity, degrading service quality for all customers. Such capacity-induced spillovers are largely overlooked in existing targeting frameworks, which assume that each customer's response to treatment is independent of others. As a result, a customer who appears profitable based on her individual response may in fact be unprofitable once the congestion she imposes on others is taken into account, thereby diminishing targeting profitability.
We propose a methodological framework for designing targeting policies that explicitly accounts for these spillovers. Our framework characterizes the conditional marginal policy effect of treatment on market-level outcomes, decomposing it into a direct effect and a congestion-mediated indirect effect. We embed this decomposition into an optimization problem that allocates treatment to maximize aggregate outcomes subject to operational constraints, yielding a tractable capacity-aware targeting policy.
Using data from a ride-hailing platform, we show that ignoring capacity constraints leads to systematic over-targeting: up to one in five customers targeted by the conventional approach imposes congestion externalities that outweigh their individual-level treatment benefit, resulting in negative net contributions to market-level profit. By correcting these targeting errors, our capacity-aware policy substantially improves campaign profit, underscoring the importance of incorporating capacity-induced spillovers into targeting decisions.
Working Papers
- "Policy-Aware Sampling: Prioritizing Consequential Customers for Optimized Targeting Policies"Conditionally Accepted at Journal of Marketing Research
Abstract
Firms often rely on randomized experiments to estimate customer-level treatment effects for targeting policies. Standard "test-then-learn" approaches typically sample customers uniformly to optimize estimation accuracy but ignore economic objectives, leading to statistically sound yet suboptimal targeting policies. We propose a policy-aware sampling strategy that directly incorporates the firm's profit-maximizing objective into the selection of which customers to sample in the experiment. Specifically, we introduce expected profit loss (EPL), a criterion that quantifies customers' learning value by measuring the financial impact of treatment effect estimation errors on targeting profitability. We show that allocating experimental resources based on EPL achieves near-optimality with theoretical guarantees, and develop a sequential sampling algorithm that prioritizes these consequential customers with highest EPLs for practical implementation. Using simulations and two empirical applications, we show that our approach yields more profitable targeting policies than existing methods. Across both applications, our approach improves targeting performance by 5% to 10% in profit terms relative to benchmark methods, and achieves comparable outcomes with up to 97% fewer experimental samples, highlighting substantial gains in both targeting effectiveness and data efficiency.
- "Dual Monetization on Creator Platforms: The Interplay Between Advertising Revenue and Brand Sponsorship"Working Paper
Abstract
Platforms featuring creator content generate revenue by showing ads against this content, and share this revenue with creators. In addition, brands pay creators upon including sponsored messages about their offerings in the content. We develop a game theory model to study the intricate strategic interactions that arise among the platform, brands, and creators in this ecosystem. A creator decides content quality and whether to include sponsored elements, where inclusion of sponsored elements implies "dual monetization" for the creator through the ad revenue shared by the platform and the payment from the brand for including sponsored elements in the content. Consumers like organic content but dislike both ads and sponsored content. We show that brand sponsorship weakly benefits the platform through a reduced revenue sharing rate because it can rely on the brand to partially compensate the creator. The creator may be worse off, even with two sources of income, because she may need to produce higher-quality content therefore incurring higher content production costs. Consumers benefit when content quality enhancements outweigh their aversion to sponsored elements. The total welfare of the ecosystem is weakly higher.