Out with the old and in with the new. In the ever-changing world of advertising technology, new trends and methods are appearing all the time. So much so, that it is hard sometimes to keep up with them.
Today we offer a quick breakdown of the two popular methods used to programmatically buy and sell digital advertising inventory: waterfall auctions and header bidding.
What Is A Waterfall Auction?
Programmatic media sales have their roots in publishers’ efforts to eliminate unsold impressions. The real-time bidding process (RTB) and other technical platforms involved in the digital display ad ecosystem originally served to let publishers quickly and easily get rid of any ad space that hadn’t already been sold through direct relationships with advertisers.
The by-product of this was the waterfall auction – also known as daisy-chaining or waterfall tags. In this process, a publisher passes its inventory from ad network to ad network in descending order of importance until all impressions are (hopefully) sold.
The ad networks are usually ranked according to the average historic yield they have produced for the publisher. This means that an ad network that where premium inventory has been sold in the past (for a higher price) will then get first chance on further impressions from the same publisher.
If, for some reason, the first network does not produce a satisfactory bid, the next ad network is called upon to bid…and so on.
Because of its relatively easy technical setup and because it helps publishers make sure that they aren’t left with any unsold ad space, waterfalling has been the method of choice for a number of years.
It is not perfect, however.
Problems with Waterfall Auctions
Waterfall auctions have some drawbacks.
Most notably, the process of “passbacks” – accepting bids from various ad networks in turn – means that there is a lack of competition for ad impressions, leading to lower overall yield for publishers.
If a certain piece of inventory was of extremely high value to an advertiser buying through an ad network that fell lower in the daisy-chain, that advertiser might be willing to bid a much higher price for it if given the opportunity earlier in the process. But without access to the auction from the beginning, they are cut out – and prices end up being lower.
Passbacks can also lead to latency. As the ad call goes to each ad network in turn, page loading times can be affected, which creates a negative user experience.
What Is Header Bidding?
Header bidding is another selling process used by publishers sell off their unsold inventory.
However,instead of the browser contacting the demand sources (e.g. ad networks and ad exchanges) one by one, header bidding brings all the demand sources together simultaneously and allows them bid on the available inventory at the same time – even before the sale of premium inventory and inventory from the publisher’s ad server. Those who return a bid in time are taken into consideration.
The biggest advantage of bringing all the buyers together to conduct a header bidding auction is that publishers can dramatically increase their yield, since:
- Publishers can bring all interested buyers to the table at the same time and look at all bids, which often results in finding a demand source wanting to purchase the inventory at a higher price than most premium buys. Also, with waterfall auctions, there may be a demand source willing to pay for more inventory, but may never get the opportunity to place a bid.
- Publishers get the real price of their inventory, not an average or estimate.
As we can see in the image above, header bidding auctions allow publishers to generate more revenue from their inventory.
Ask us anything header bidding and waterfalling
Downsides of Header Bidding
Although it would seem that header bidding solves many of the problems with waterfall auctions, it is not without its drawbacks.
For one thing, it is technically more complicated to implement with code to connect a publisher’s page to each and every SSP/ad exchange/ad network needing to be added into the page in question.
For another, header bidding can produce latency problems too, as each tag placed in the header has the potential to slow down the page load time.
And then there is the issue of duplicate bidding. An advertiser who is plugged into several ad networks could end up bidding against itself without knowing it, which in turn can negatively impact the performance, including the amount of data exchanged, of the ecosystem.