Authors: Panagiotis Papadopoulos, Nicolas Kourtellis, Pablo Rodriguez Rodriguez, Nikolaos Laoutaris
Abstract: Online advertising is progressively moving towards a programmatic model in which ads are matched to actual interests of individuals collected as they browse the web. Leing the huge debate around privacy aside, a very important question in this area, for which little is known, is: How much do advertisers pay to reach an individual? In this study, we develop a first of its kind methodology for computing exactly that – the price paid for a web user by the ad ecosystem – and we do that in real time. Our approach is based on tapping on the Real Time Bidding (RTB) protocol to collect cleartext and encrypted prices for winning bids paid by advertisers in order to place targeted ads. Our main technical contribution is a method for tallying winning bids even when they are encrypted. We achieve this by training a model using as ground truth prices obtained by running our own “probe” ad-campaigns. We design our methodology through a browser extension and a back-end server that provides it with fresh models for encrypted bids. We validate our methodology using a one year long trace of 1600 mobile users and demonstrate that it can estimate a user’s advertising worth with more than 82% accuracy.
- We propose the first to our knowledge holistic methodology to calculate the overall cost of a user for the RTB ad ecosystem, using both encrypted and cleartext price notifications from RTB-based auctions.
- We study the feasibility and efficiency of our proposed method by analyzing a year-long weblog of 1600 real mobile users. Additionally, we design and perform an a affordable (a few hundred dollars cost) 2-phase real world ad-campaign targeting ad-exchanges delivering cleartext and encrypted prices in order to enhance the real-users’ extracted prices. We show that even with a handful of features extracted from the ad-campaign, our methodology achieves an accuracy > 82%. The resulting ARPU is ∼55% higher than that computed based on cleartext RTB prices alone. Our findings challenge the related work’s basic assumption that encrypted and plain text prices are similar (we found encrypted prices to be ∼1.7× higher). Finally, we validate our methodology by comparing our average estimated user cost with the reported per user revenue of major advertising companies.
- Using lessons from the study, we design a system where the users, by using a Chrome browser extension, can estimate in real-time, in a privacy-preserving fashion on the client side, the overall cost advertisers pay for them based on their exposed personal information. In addition, they can also contribute anonymously their impression charge prices to a centralized platform for further research.
Read the entire paper here.