New Advertising – Who Do VCG Algorithms Benefit?
https://www.wired.com/2015/09/facebook-doesnt-make-much-money-couldon-purpose/
In the article, it describes the two most popular and fundamental methods behind the advertisement placing algorithms on the internet. These models are highlighted by using the two most popular internet destinations today, Google and Facebook. The first, older form of ad placement technique is used by Google to fill out their search engine platform with ads. The tech giant uses a generalized second price auction in which advertisers will place bids to Google in order to get an ad placed for specific search terms with the winning bidder paying the price of the second highest bid. This creates a dominant strategy to bid truthfully at the value you place on the advertisement and maximizes revenue by selecting only advertisements that are worth the highest payments from their creators. Facebook revolutionized ad-placements by utilizing the Vickrey-Clarke-Groves auction system within their algorithm. This method does not select advertisements based on the amount that advertisers are willing to pay for them in favor of prioritizing the net benefit of the whole system. This is supposed to select ads that are more relevant and ideal for the viewers and thus make the ads selected more effective, increasing the amount bidders would pay for advertisements over the long term.
VCG auction systems have revolutionized the way advertisers think about ad selection and creation by eliminating ways to game the system and thereby raising the overall benefit and effectiveness of the system. VCG algorithms calculate the maximum benefit of the overall system by calculating the combinations of buyers and sellers (ads and the spaces where they can exist on the internet) that produces the maximum social benefit and then charging the advertiser the VCG price. These are highly intricate networks that can only function in an auction where the seller first receives all the bids from the buyers who in this case are advertisers. For Facebook, this has proven extremely successful and could eventually become the model used by everyone, especially with access to big data becoming more prevalent and useful for targeting consumers. With more data on consumers, the easier it will be for VCG algorithms to function and correctly predict the changes in valuations so as to maximize the social benefit and value.