Measuring the Return to Online Advertising:Estimation and Inference of Endogenous Treatment EffectsShakeeb Khan1, Denis Nekipelov2, Justin Rao3AbstractIn this paper we aim to conduct inference on the “lift” effect generated by an online advertisementdisplay: specifically we want to analyze if the presence of the brand ad among theadvertisements on the page increases the overall number of consumer clicks on that page.A distinctive feature of online advertising is that the ad displays are highly targeted- theadvertising platform evaluates the (unconditional) probability of each consumer clicking on agiven ad which leads to a higher probability of displaying the ads that have a higher a prioriestimated probability of click. As a result, inferring the causal effect of the ad display onthe page clicks by a given consumer from typical observational data is difficult. To addressthis we use the large scale of our dataset and propose a multi-step estimator that focuseson the tails of the consumer distribution to estimate the true causal effect of an ad display.This “identification at infinity ” (Chamberlain (1986)) approach alleviates the need for independentexperimental randomization but results in nonstandard asymptotics. To validateour estimates, we use a set of large scale randomized controlled experiments that Microsofthas run on its advertising platform. Our dataset has a large number of observations and alarge number of variables and we employ LASSO to perform variable selection. Our nonexperimentalestimates turn out to be quite close to the results of the randomized controlledtrials.JEL Classification: C14, C31, C55, C90, M37.Keywords: Endogenous treatment effects, randomized control trials, online advertising, lifteffect.
Measuring the Return to Online Advertising_Estimation and Inference of Endogenou.pdf
(661.13 KB, 需要: 3 个论坛币)