Each month, Basis Technologies’ Programmatic 101 series tackles a different facet of programmatic advertising—from best practices for buyers, to competitors in the space, to trends you should know.
One of the most complex aspects of programmatic buying is the ability to decide how much you’re willing to pay for an impression. As the advertising industry has evolved, so too have the bidding models and strategies available. Below, we dive into the evolution of bidding models, from waterfall auctions and header bidding to first- and second-price auction models and bid shading.
Before we get into the different bidding methodologies, let’s get clear on the meaning of real-time bidding, or RTB. Contrary to common misconception, RTB is not synonymous with programmatic!
RTB refers to a buying method used to purchase a single impression via an auction-based system in real-time. Programmatic, on the other hand, is a broad term that represents the automated technology infrastructure that enables media buying.
So, you can use RTB to buy inventory programmatically, but there are other ways to buy inventory that are not in real-time, such as programmatic direct and preferred deals. In other words, RTB is always programmatic, but programmatic encompasses more than just RTB.
Now that we’re clear on what RTB entails, let’s talk about the various models under the RTB header.
When RTB was first introduced, inventory was sold to one ad network at a time. The term ‘waterfall auction’ describes the norm for bidding during this time.
In a waterfall auction, publishers reach out to ad networks one by one, giving priority to larger networks and networks that have yielded more revenue for a publisher in the past. This means that major players like Google Ad Network receive more exposure to auctions, even if there are other networks willing to pay a higher cost.
Header bidding was introduced to address the inequitable nature of waterfall auctions. In the header bidding model, all buyers bid at the same time. This model provides an equal playing field for all buyers despite their size or revenue yield. Today, header bidding has almost achieved the level of an industry standard.
First-price and second-price bidding models are yet another variable to be considered in RTB. In the first-price bidding model, the buyer pays the exact amount they placed for their bid. For example, let’s say Advertiser A bid $4.00, Advertiser B bid $3.00, and Advertiser C bid $3.75. Advertiser A would win the auction because their bid was the highest, and they would pay $4.00. Since buyers have no idea what their competitors are bidding, this method leads to higher bids.
In the second-price bidding model, the buyer pays $0.01 more than the second-highest bid. Using the example above, in a second-price auction, Advertiser A would pay $3.76, or $0.01 more than the second-highest bid.
While second-price bidding has historically been the standard, giants like Google Ad Manager have moved to the first-price model in recent years.
For buyers to remain competitive in a world of first-price auctions, bid shading has emerged as an automated method for bidders to better determine bid amounts. Bid shading algorithms analyze historical bid data, looking at factors like site, ad size, exchange, and competitive dynamics to inform bid recommendations. Bid shading is so nuanced that it can help predict bid prices by tactic, which allows publishers to receive the highest bid while ensuring buyers keep their bids as cost-effective as possible.
To learn more about the benefits of bid shading, check out Why Bid Shading Should Be Part of Your Programmatic Strategy.
There you have it: from waterfall auctions to the introduction of bid shading, RTB has seen many evolutions since its beginning.
Hungry for more? Check out Basis’ Programmatic Advertising Essentials Certification to gain a fundamental understanding of the programmatic landscape.
Plus, keep your eyes peeled for next month’s installment of Programmatic 101 to learn more about Private Marketplace deals, another form of real-time bidding!