Ad Auction Design and User Experience

Abstract

Abstract. When users click on poor quality advertisements, there is a hidden cost to the search engine due to the user dissatisfaction (for instance, users are less likely to click on ads in the future). We describe how to incorporate hidden costs into the GSP auction for internet ads such that it is in an advertiser's self interest to create a user experience that maximizes efficiency.

Correcting Audience Externalities in Television Advertising

“…A second approach to handling audience externalities was developed by Abrams and Schwarz (2008), also in the context of search advertising. Under their model, each advertiser chooses an offering (such as a landing page); different offerings have different tradeoffs between nuisance to the consumer and profit to the advertiser.…”

“…The key modeling distinction is that the nuisance to the consumer directly translates into a cost to the search engine, in that a higher nuisance reduces consumers' propensity to click on ads in the future, and thus the search engine's future revenue. The main technical contribution of Abrams and Schwarz (2008) is an extension of the Generalized Second-Price (GSP) auction of Edelman et al (2007) and Varian (2007) to explicitly charge advertisers for nuisance costs and thereby provide incentive for each advertiser to choose the offering that maximizes the joint welfare of the advertiser and search engine. Stourm and Bax (2013) showed that when ads with higher nuisance costs also provide more value to their advertisers, the GSP extension of Abrams and Schwarz (2008) will improve the advertisers' collective welfare when they are sufficiently numerous.…”

“…The main technical contribution of Abrams and Schwarz (2008) is an extension of the Generalized Second-Price (GSP) auction of Edelman et al (2007) and Varian (2007) to explicitly charge advertisers for nuisance costs and thereby provide incentive for each advertiser to choose the offering that maximizes the joint welfare of the advertiser and search engine. Stourm and Bax (2013) showed that when ads with higher nuisance costs also provide more value to their advertisers, the GSP extension of Abrams and Schwarz (2008) will improve the advertisers' collective welfare when they are sufficiently numerous. Similar to Aggarwal et al (2008) and Kempe and Mahdian (2008), these papers assume homogeneity in how audiences respond to ads.…”

Optimal auction design in two‐sided markets

“…The results from Edelman and Schwarz () imply that, when the advertisers' types are one‐dimensional, the Generalized English auction (with appropriate reserve prices) considered by Athey and Ellison () is the profit‐maximizing mechanism. In a different setting, Abrams and Schwarz () allow advertisers to choose the “quality” of their landing page while competing in a GSP auction. Abrams and Schwarz introduce a Pigovian tax in the GSP framework that aligns the advertisers' and the platform's interests regarding the provision of quality.…”

Automated creation and optimization of online advertising campaigns

“…Public taxonomy and concept hierarchy. Chen et al [18] introduce the use of taxonomies available on the web (like DMOZ 3 ) to build a tree of concepts.…”