Bid Shading
https://www.adweek.com/programmatic/bid-shading-is-covering-for-inefficient-programmatic-algorithms/
In the marketing industry, programmatic advertising has become the norm. More recently, the industry has seen a shift to first-price auctions from second-price auctions where winners pay the amount that they bid rather than the amount of the second highest bidder. This shift has caused marketers to experience a sudden increase in costs, causing them to turn to different methods to mitigate these losses. One method that has proliferated is bid shading. Bid shading uses an algorithm that alters the bidder’s price, lowering it based on a variety of factors including ad size, ad location, and the exchange.
However, bid shading has merely caused an illusion of successful, efficient programmatic bidding when in reality, it is anything but that. This is because the algorithm that bid shading depends on doesn’t consider the expected cost of the inventory before placing the bid. Therefore, it is only useful if and only if the bid placed is on an effective product, it doesn’t help choose the most efficient product to bid on. One solution to this problem is the development of more accurate algorithms. With the use of machine learning and artificial intelligence, it is possible to develop programmatic bidding strategies that also encompass where brands should spend their money on and what consumers of their ads are more likely to convert.
This article relates to the concept of first-price and second-price auctions that was introduced in class. With second-price auctions bidders are more likely to bid their true value, while with first-price auctions, bidders are more likely to bid slightly less than their true value in order to ensure they make a profit. This is what has occurred with bid shading where the algorithm lowers the bid price of the buyer in order to secure a profit. According to the article, many companies are turning to first-price auctions to increase transparency since bidders will be paying exactly what they are bidding. However, with bid shading and the transparency that it inherently lacks, these efforts are useless.