Facebook: Optimizing Total Value Over Monetary Value
Sources: https://www.facebook.com/business/help/430291176997542?helpref=faq_content
While Google is considered to be the largest ad platform for industries and businesses across the spectrum, Facebook is perhaps the most powerful social media platform that exists today that makes it an ideal tool for companies to advertise themselves. As a result, Facebook too plays a significant role in selling ads through bidding and other methods.
Let’s break down how Facebooks holds ad auctions. When Facebook shows an ad, it strives to balance out the following two objectives:
- Creating value for the advertisers
- Providing positive and relevant experiences for Facebook users
That is, the winning ad is the one that best represents both objectives as equally as possible and provides the most total value rather than the one that always has the highest bid as in a traditional auction. Thus, Facebook doesn’t just consider the advertiser’s best interest but strives to maximize the value of both parties affected by an ad posted on the news feed.
So how does Facebook determine which ad will provide the most total value? Three factors determine the winner of a bid:
- Advertiser bid
- Estimated action rates
- Ad quality and relevance
Clearly, advertiser bid refers to the monetary value an advertiser bids. The dominant strategy for an advertiser is to bid his/her true value or higher. According to the source, the type of auction Facebook runs can somewhat be compared to a second-price auction. However, rather than paying the second highest bid, the winner pays the minimum amount needed to win the auction, which usually means the winner pays less than what he/she bid. This means bidding at least the true value suffices and that bidding less than the true value is not an advantageous strategy for an advertiser.
Estimated action rates refer to how likely someone will commit a certain action such as clicking a link, liking a page, or viewing a page. These estimates are based on the ad’s past performance and on the previous actions of the audience members the ad is targeted at. With estimated action rates, Facebook can see how relevant an ad is to a given audience’s interests and how likely someone is to commit the type of action the advertiser wants its audience to take, optimizing the total value Facebook strives to fulfill when auctioning off ad sets.
Finally, ad quality and relevance are rather straightforward. Ad quality and relevance play a large role in determining the total value an ad provides and is usually determined by something called a relevance score, which is comparable to PageRank values. Positive and negative feedback to an ad will increase or decrease its total value respectively, and relevance between a user’s history and an ad (i.e. how often the user viewed or clicked on ads provided by a certain advertiser) can affect the ad’s total value.
In class, we investigated Google’s own form of auctioning ads, GSP. With Facebook, we see that the social media giant has its own system of auctioning ads as well that too deviates from the traditional auction types (first-price auctions, second-price auctions, etc) we discussed earlier in the semester. Even more, while Google runs auctions such that the bidder with the highest monetary value wins the best ad slot, Facebook diverges even more from convention by optimizing total value (the monetary value gained by the advertiser + the personal experiences gained by users). The bidder with the highest bid can still win. However, if quality and relevance of one ad is much better than that of an ad with the highest bid, then the first of the two will win. Unlike most other auctions, Facebook doesn’t care just for businesses and the monetary value they gain from advertising on Facebook, but, as a social media platform, it also demonstrates a great amount concern for ensuring positive and quality user experience for the community.