Impact of Disabling Third-Party Cookies on Publisher Revenue

This technical report empirically quantifies the effect of disabling third-party cookies on the programmatic ad revenue of web publishers. The study involved a randomized controlled experiment conducted on publishers using Google Ad Manager's serving system. The experiment disabled access to cookies for a small fraction of randomly selected users, leading to a significant decrease in average revenue for the top 500 global publishers. Specifically, the average revenue in the treatment group decreased by 52%, with a median decline of 64%. The methodology included a robust experimental infrastructure for running A/B tests, comparing revenue from users who saw non-personalized ads to those who received personalized ads. The results indicated that while some publishers experienced minimal revenue loss, the majority faced declines of 50% or more. The report also discusses related work and the challenges of drawing causal conclusions from observational studies, highlighting the need for further research on second-order effects of cookie removal.