Header bidding turns the scarcity that has so far fuelled desire for video on its head. Instead of keeping buyers keen enough to pay premium prices by limiting supply to direct deals, it opens the floodgates to all demand and uses competitive bidding to drive yield sky high.
At least, that’s the theory, and there’s no denying that the header bidding principle has a distinct allure. When it first emerged on the digital advertising scene last year, its potential to replace the restrictive waterfall model with a real-time, multi-buyer auction that could take place before calling the ad server was hailed as revolutionary. Today confidence in its abilities is so great that nearly half of digital publishers are using the technology.
But for all the industry enthusiasm, does header bidding have what it takes to dominate the future of automated ad trading and, crucially, is it up to the challenge of digital video?
Heavy tags are still hindering lift-off
No matter how advanced a system may be, if the weight of usage is greater than its capacity, something has got to give. And for header bidding, that something is time. Aside from the difficulties of installing its intricate scripts, latency is the biggest issue with standard header bidding. Embedding header tags within pages that simultaneously call a host of ad exchanges may enable publishers to access a greater range of demand, but it can also weigh down the page and drain user resources, resulting in frustratingly slow loading times.
While such teething problems aren’t unusual for new technologies, a glitch that negatively impacts the user experience is not good news for publishers or advertisers when other content is only a click away. If header bidding is to achieve lift-off, more work is needed.
Video brings its own layer of complexity
Maintaining a fast, efficient trading process with a simple format like display is tricky. Add video to the mix and the task becomes a mammoth one. Ad units not only come in different sizes that take longer to load, from in-stream pods to autoplay, but they are also managed by multiple siloed technology platforms and must adapt to the specifications of each mobile device. What’s more, existing advertising standards, namely the Video Player Ad-Serving Interface Definition (VPAID), do not provide a clear view of exact bid price
So, the header bidding system that works well for display — whereby lightweight creative is cached in the header ready for demand to be called instantly — cannot run as smoothly with larger video creative that uses VPAID guidelines and in doing so, disrupts cache logic. The result is an experimental and inconsistent approach with significant latency on the client side, which impedes vital user engagement and interaction with video ads.
Finally, we come to the most considerable challenge for video header bidding — the fact it was created as an alternative to Google’s non-transparent DoubleClick Ad Exchange (ADX). Intended to fix ADX’s inability to facilitate real-time external demand, header bidding did provide a way to avoid basing bids on estimates of what outside exchanges might offer. But it was only a transitional solution and now Google has released its own answer, dynamic allocation, the need for header bidding could be about to subside. Much like Flash was all-but eradicated when it failed to function in all markets and replaced with HTML5, header bidding may loose its current hold on digital advertising if Google decides to reassert itself.
Can these problems be solved?
In an industry driven by innovation, attempts to address these issues by upgrading header bidding were inevitable. The most widely adopted include; a data-based system that tries to make tags more manageable by inserting weighty but adjustable wrappers in page headers, adding mediation layers directly into video players, and amassing several ad calls — an approach that frequently causes slow page loads or time-outs and irritates users.
Particular focus has fallen on out-stream video — ad units that appear within text, not video content — as latency here can severely impact publishers. Some upgrades have sought to reduce latency by calling ad exchanges from the server, rather than the browser, and thereby lighten the load on user connections. Yet this has created problems of its own, with less transparency, limited opportunities for buyers and more fraud, as well as being in contrast to VPAID regulations, which dictate that demand must function on the client-side.
So what’s the future?
Occupants of the digital ecosystem are finding that users increasingly want fluid advertising experiences that align with their connected lifestyle — an expectation that fragmentation makes it hard to deliver on. In the future we are therefore likely to see greater consolidation, with tech providers working side by side to create the tools needed for cohesive marketing.
In such a landscape the fate of header bidding is uncertain. If Google’s dynamic allocation is successful, it’s probable the technology will be eclipsed by a better solution made by media giants with the resources required to stay ahead. If not, header bidding might get a revamp at the hands of third-party developers, who could — with the right tweaks — create a tool that outpaces conventional header bidding and Google. And if the industry does choose to embrace integration or a new wonder technology emerges, it’s possible that header bidding may disappear completely.
Whatever the future has in store for header bidding, one thing is clear: the technology is not, at present, up to the challenge of video. Complex, diverse and unwieldy, video formats are simply too much when added to its already difficult balancing act of multiple buyers, sellers, exchanges and servers. Given the odds stacked against it, my bet is that wherever header bidding does end up, it won’t be as the dominant force in video.
(As published on VideoAdNews)
January 18, 2017
January 17, 2017
January 05, 2017
December 28, 2016
December 20, 2016
December 08, 2016