Ghost Ads: Improving the Economics of Measuring Ad Effectiveness

To measure the effects of advertising, marketers must know how consumers would behave had they not seen the ads. We develop a methodology we call ‘Ghost Ads,’ which facilitates this comparison by identifying the control-group counterparts of the exposed consumers in a randomized experiment. We show that, relative to Public Service Announcement (PSA) and Intent-to-Treat A/B tests, ‘Ghost Ads’ can reduce the cost of experimentation, improve measurement precision, and work with modern ad platforms that optimize ad delivery in real-time. We also describe a variant ‘Predicted Ghost Ads’ methodology that is compatible with online display advertising platforms; our implementation records more than 100 million predicted ghost ads per day. We demonstrate the methodology with an online retailer’s display retargeting campaign, for which a PSA test would be severely biased. We show novel evidence that retargeting can work as the ads lifted website visits by 17% and purchases by 11%. Compared to Intent-to-Treat or PSA experiments, advertisers can measure ad lift just as precisely while spending at least an order of magnitude less. Johnson: Simon Business School, University of Rochester, . Lewis: Netflix, . Nubbemeyer: Google, . We thank seminar participants at Kellogg, Stanford GSB, UCSD Rady, Abdelhamid Abdou, David Broockman, Hubert Chen, Mitch Lovett, Preston McAfee, John Pau, David Reiley, Robert Saliba, Kathryn Shih, Robert Snedegar, Hal Varian, Ken Wilbur, and many Google employees and advertisers for contributing to the success of this project.