Optimal Microtargeting of Advertising
将最优控制理论从宏观预算分配扩展到微观客户层面，通过估计个体广告响应参数，确定每位顾客在各媒介的最优曝光次数，实证表明该方法可将利润提升150%至183%。
Historically, advertising allocation decisions have operated mostly at a macro level, comprising determination of the total budget followed by apportionment among media channels. Owing to the rapid and sustained rise of digital media, an additional decision now operates at a micro level, namely, which specific customers to target with advertising. In the macro case, optimal control theory provides a powerful framework for firm profit maximization, specifically allowing for ad response, cost per medium, and discount rate, all in the presence of multiple competing brands. However, optimal control theory has never been applied to the situation of microtargeting individual customers. Consequently, in this study the author shows how optimal control theory can be adapted for application to individual customers. The author estimates a multinomial logit model with individual-specific advertising response parameters. In turn, these parameters are used to determine the optimal number of exposures each customer should receive for each advertising medium. An empirical example demonstrates that using the optimal microtargeting method improves firm profits over existing ad scheduling methods by between 150% and 183%.