Why Digital Display Advertising Moved from Second-Price to First-Price Auctions
In digital display advertising today, most auctions operate as first-price auctions. This represents a significant shift from earlier market designs and stands in contrast to what we still see in search advertising and in many traditional auction settings. The move has been controversial, with some practitioners arguing for a return to second-price auctions. To understand why the industry changed direction, it helps to clarify how the two auction types work, why economists historically favored second-price auctions, and why those advantages eroded in real-world display advertising markets.
First-Price and Second-Price Auctions Explained
In a first-price auction, the highest bidder wins and pays exactly what they bid. If one buyer bids $5, another bids $4, and a third bids $3, the $5 bidder wins and pays $5. The mechanism is simple and transparent.
In a second-price auction, the highest bidder still wins, but they pay the price of the second-highest bid, often plus a nominal increment. Using the same bids, the $5 bidder would win but pay $4 (or $4.01). The key difference is that the winning bidder does not pay their own bid.
Why Economists Prefer Second-Price Auctions
Economists tend to favor second-price auctions because they encourage truthful bidding. In theory, bidders are incentivized to bid the true value they assign to the item, since they expect to pay less than their bid if they win. In a first-price auction, by contrast, bidders have an incentive to “shade” their bids below their true valuation in order to preserve surplus. Second-price auctions simplify bidding strategies and, under ideal conditions, lead to efficient outcomes.
This logic explains why second-price auctions have been used successfully in controlled, centralized environments such as search advertising, where a single platform runs the auction and enforces consistent rules.
Why Second-Price Auctions Broke Down in Display Advertising
Display advertising markets introduced complexities that undermined the assumptions required for second-price auctions to work cleanly. One major issue was the rise of what became known as “modified” second-price auctions. In practice, this label could mean many different things. Publishers or intermediaries might introduce price floors that effectively act as bids. These floors could be static or dynamic and were not always visible to buyers. In other cases, markups larger than a nominal increment were added. Some systems even inserted synthetic or “dummy” bids that functioned like hidden floors. While these mechanisms may have been disclosed in theory, they were often opaque in practice, making it difficult for buyers to understand the true clearing price.
The result was a loss of trust. Buyers could no longer be confident that a second-price auction was actually operating as advertised. From the buyer’s perspective, they might submit a bid believing they would pay the second-highest bid, only to discover they were paying much more due to hidden adjustments.
The Problem of Sequential Auctions and Self-Competition
A second, deeper issue emerged from the fragmented supply chain in display advertising. Instead of a single auction, impressions often passed through multiple intermediaries, each potentially running its own auction. Buyers could end up bidding into several parallel or sequential auctions for the same impression without full visibility into the process.
In these scenarios, a buyer could unknowingly compete against themselves. Their own bids, submitted through different paths, could act as de facto price floors in downstream auctions. In effect, what was theoretically a second-price auction would collapse into a first-price outcome, with the buyer paying close to their own bid rather than the next-best competitor’s bid. This dynamic contributed to the rise of concepts like supply path optimization, as buyers tried to reduce redundant paths and avoid self-competition.
Why First-Price Auctions Became the Default
Given these realities, the theoretical advantages of second-price auctions no longer held. Transparency suffered, trust eroded, and the fragmented auction environment made true second-price mechanics difficult to enforce. In contrast, first-price auctions, while requiring more sophisticated bidding strategies, are conceptually straightforward. Buyers know that if they win, they pay what they bid.
From the publisher side, there was an additional incentive. When the market initially shifted to first-price auctions, many buyers did not immediately adjust their bidding algorithms. As a result, publishers often saw a short-term lift in yield, as bids reflected legacy second-price assumptions. While this effect diminished over time as buyers adapted, it reinforced publisher acceptance of the new model.
Why a Return to Second-Price Auctions Is Unlikely
For a second-price auction to function as intended, it requires a single, trusted marketplace with clear and enforceable rules. That condition exists in some environments, such as search, but it does not describe the modern display advertising ecosystem. The combination of multiple intermediaries, opaque adjustments, and fragmented auctions makes it extremely difficult to restore the trust and simplicity required for second-price auctions to work at scale.
As a result, despite ongoing debate, first-price auctions have become the dominant model in display advertising, and there is little indication that the industry will revert anytime soon.