In the formative years of programmatic advertising, second-price auctions were the industry standard—a crucial component in helping build the online ad marketplace as we know it today. Much like eBay, ad exchanges saw second-price auctions as a better, more accurate valuation of publishers’ inventory.
This all changed back in 2017, though, as the major exchanges began either rolling out or experimenting with first-price auctions, culminating in Google joining the pack in 2019. It was an industry-wide move largely dictated by increasing calls for greater transparency into the bidding process and reduced operational complexity. The programmatic ecosystem was essentially becoming so difficult to navigate within a second-price framework that the market needed a switch.
One of the manifestations of this evolution was bid shading—an AI-powered optimization tactic designed to help media buyers reduce wasted ad spend in the new auction dynamic. Here, we break down some of the lingo surrounding bid shading and explore how it’s applied in the digital advertising industry.
Second-price auctions refer to a model in which the buyer pays just $0.01 more than the second-highest bid on an impression (think the eBay model). For example, if two buyers bid $10 and $5, respectively, then the buyer who bid $10 will win the impression—but they’ll only pay $5.01. In other words, it is the second-highest bidder that determines the clearing price… in theory, that is. The problem with this bidding framework (and why it has largely been shunned by the digital advertising industry) is that some ad exchanges aren’t truly operating on the second-price framework. Each exchange has its own variation on the model with nuances that don’t offer full transparency (think price floors, hidden supply-side fees, and advertising subsidies).
In a first-price framework, the auction will clear at the winning bid price outright. Meaning: if a buyer bids at $10 and the next highest bid is $5, the winning buyer will pay the full $10. While this represents a more attractive model for publishers, advertisers may find themselves overspending and paying an increased average cost per 1,000 impressions (CPM). As such, advertisers need new tools that empower them to bid more effectively and intelligently.
Enter bid shading.
A compromise between the two models, bid shading is an optimization tactic available in most enterprise demand side platforms (DSPs). It works by analyzing historical bid data, then automatically forecasting a winning bid that is lower than the default bid (though likely more than just $0.01 above the next highest bid). For instance, if buyer A bids $10 and buyer B bids $5, buyer A might pay $7.50 rather than the full original value of their bid.
Before this tactic emerged, advertisers bidding in first-price auctions needed to either bid high and potentially overspend on an impression or bid conservatively and potentially lose the impression. By activating bid shading, media buyers can unlock two critical things:
Easy: any agency or brand (regardless of industry) looking to gain more efficient CPMs.
Yes, bid shading works with a range of other optimization tactics, including algorithmic optimization (AO), machine learning optimization (MLO), and group budget optimization (GBO). Bid shading aims to decrease the bid price without changing what any of those tactics are optimizing towards. This keeps the probability of winning the auction high enough that it will not compromise pacing and target budgets. Additionally, if advertisers have manual optimizations on domains and placements, the manual bid price applies first, and then bid shading takes effect.
Supply path optimization (SPO) is another name for the algorithms that DSPs use to make sure they're bidding on the most relevant, highest-quality, and most valuable inventory available from supply-side platforms (SSPs). When it comes to the economic facets of of SPO, bid shading can be a key tool, helping DSPs ensure they are paying the lowest possible amount on any given bid and, potentially, eliminating SSPs that don't provide second-price auctions—all to ensure optimal value.
Yes, digital advertisers can use bid shading with PMP deals.
Bid shading generally leads to a more efficient CPM, delivering more impressions and potentially increasing ad serving fees and impact pricing. As such, buffering ad serving fees is recommended.
No, though advertisers that use a default CPM of $1 or greater are more likely to see positive results.
At the most rudimentary level, bid shading algorithms analyze past placement clearing prices, compare them to what a buyer is willing to pay, and then make their best estimation at the lowest price the buyer could submit to win the auction. That said, there are many other factors that could inform the logic—including ad size, floor prices, domain information, time of day, win rates, placement on page, and user data—and every DSP goes about determining the final bid in a different way.
As the programmatic advertising landscape continues to evolve, so do the problems that come with it. The industry’s broad shift from second- to first-price auctions was one such problem that media buyers needed to contend with, and DSPs had to move quickly to assuage the price surges that surfaced from the move. Bid shading was the result.
Today, the tactic is widely adopted, and it represents a great example of just how important automation is in modern advertising. It helps advertisers save time, reduce menial manual tasks, and optimize eCPM leads to deliver better results.
Want to learn more about advertising automation? Check out our guide to see why automation is essential to the future success and long-term growth of the media buying industry.