For all of digital advertising's countless acronyms and obscure technologies, no piece of industry jargon enjoyed a greater surge in popularity this past year than “header bidding." Over the course of 2016, header bidding went from a monetization tool used by a handful of savvy publishers to a phrase that was on the lips of just about everyone bearing a name badge at conferences and cocktail parties. By August, a survey from AOL would reveal that roughly half of U.S. and U.K. publishers had already implemented it.
All of this is to say that by now, you're probably familiar with header bidding -- or at the very least, you know to nod approvingly when someone else brings it up in conversation. But what you might not know is how it actually works. Indeed, for all the chatter about header bidding, the tool remains a great source of confusion for many publishers.
Fortunately, we're here to help. Let's take a quick look at where header bidding came from, why it became popular, and where it's going next. By the time you're done reading, you'll have all the information you need to sound smart when ad-tech's hottest buzzword pops up at your next meeting.
A brief history of header bidding
At its core, header bidding is a tool that helps publishers earn higher revenues from their inventory by allowing multiple ad networks and exchanges to bid on an impression at the same time. If that sounds somewhat opaque, it's probably helpful to understand the method publishers were using to sell their inventory before header bidding came around — a process known as “waterfalling."
In a waterfall auction, publishers create a hierarchy that determines the order by which their demand partners can bid on programmatic inventory, with each demand source offering their proposed price one at a time. If the highest-ranked bidder (whether an exchange or network) produces a bid above a certain price floor, it wins the auction. If not, the impression is offered to the next highest-ranked partner, who can either meet the price floor or send the impression further down the ladder. As an unfortunate side effect, publishers leave money on the table every time a high-ranking partner wins an impression at a price lower than what someone at the back of the line would have paid for it.
Header bidding was created to solve this problem by allowing all of a publisher's programmatic buying partners to compete on equal footing for inventory. Though the technology was first introduced in 2009, it has only became popular over the past two years — a testament to publishers' ever-improving technical knowhow. The way the tool works is that publishers add a string of code, known as a “wrapper," to the header of their web pages. The wrapper then submits simultaneous bid requests to demand partners before the ad server is called.
Since every potential buyer participates at the same time, publishers can sell each impression to the highest bidder, rather than to the one that just happened to have the best spot in the waterfall. In addition, header bidding proponents say that the increased competition for publisher inventory drives up prices over time, as it encourages advertisers to offer higher CPMs to beat out the other bidders.
A video evolution is on the horizon—or is it?
Though it has become extremely common for publishers to use header bidding to squeeze every last cent out of their display inventory, header bidding for video is only just gaining steam. Largely, this is due to the greater technical complexity of video implementation, rooted in the fact that video players don't actually have headers.
At first, technology companies began developing solutions that held the header bidding auction inside the video player itself. But because video ad units come in larger file sizes than display units, publishers are concerned that holding the auction inside the player may contribute to slow load times that could hurt the user experience. More recently, publishers have started experimenting with server-to-server connections, which function a lot like a regular SSP. Rather than calling every demand partner from inside the video player, the player makes a single call to an outside server, which then calls the demand partners. As a result, this shift to server-to-server setups has reduced site load times and eliminated the need for wrappers in video header bidding. So, is what some people are calling "header bidding for video" really something that's already been solved, at least pertaining to yield management?
Given the massive growth of video and its increasing share of the publisher revenue mix, one can only assume that more content owners will explore video header bidding solutions. In fact, several major publishers are already adopting server-to-server video header bidding products. As with any new technology, video publishers will need to develop a deeper understanding of header bidding in order to determine whether it's right for them. We hope this brief background will serve as a useful first step in this important education process. Stay tuned for more as the conversation around header bidding for video continues!
As a leading platform for sellers of video advertising, we're always working to help publishers and content owners solve their most pressing challenges. Here's what we're doing to stop ads from buffering.