Wait a minute – isn’t that what an RTB auction does? Not exactly. In a real-world auction, the universe of demand is all sitting in the same room, and all bid concurrently for the item up for sale – thus ensuring the item is sold at the highest possible price.
In the programmatic world, due to the fragmentation of demand, the bidding happens sequentially. The first bidder group walks into the room, buys whatever they want, and walks out. The next bidder group is then invited into the room, buys what they want, and so on – the famous daisy chain or waterfall. The value of the impression isn’t necessarily maximized, as a subsequent bidder may have been willing to pay more for the impression than an earlier bidder, but they never got the chance. This sequencing creates an additional issue when the item up for sell is a user’s eyeballs, which can move on before the second bidder gets into the room and cause impressions to be lost and/or never monetized.
Advance bidding – which implies bidding that happens prior to the clearing process run in the ad server – provides publishers a means of ensuring each impression is seen by multiple buyers, and ultimately sold to the highest bidder. In essence, it blows up the daisy chain, allowing publishers to see how different programmatic partners perform when given equal and concurrent access to their inventory.
So how exactly does advance bidding – also sometimes confusingly described as a "tagless" solution or header-tag integration – work? It’s accomplished by placing a demand partner’s tag into a Web page's code, therefore enabling that partner to place a bid before the traditional ad-server unit is called. That bid is carried into the ad server’s auction, creating more competition in determining the winner. Since the bids are all evaluated concurrently, versus sequentially, the real highest bidder wins.
Ironically, the march toward this solution has been led mostly by buyers rather than publishers. These buyers want direct-to-publisher access and a way to get a first or exclusive look at all the available inventory. Publishers can use this to their advantage; when multiple demand partners have first-look access to the inventory, it creates more competition (as a true bidding war) for the pricing of that inventory. This also ultimately benefits buyers, since those who differentiate themselves through their performance are likely to be given preference in the future.
This all sounds great. But what’s the catch? Advance bidding can be cumbersome and time-consuming. The initial integration of partners – who all work differently – takes time. There's also concern over data leakage and user privacy, as the browser exposes everything to the bidder and could allow any partner to drop a cookie (even if the partner doesn't buy or even bid on the inventory). The number of partners publishers can integrate is limited, as partners' response times could increase page-load times. There is also a limit to how many disparate browser-to-server conversations can be supported simultaneously.
While not a perfect solution, by increasing competition, publishers can realize the value of more partnerships and the unique demand they each provide. So with further refinement, advance bidding could turn demand fragmentation from a limitation to a publisher’s advantage – thus solving a big piece of the programmatic puzzle.