Header Bidding Doesn’t Work For Video

On TV And Video” is a column exploring opportunities and challenges in advanced TV and video.

Today’s column is written by Scott Braley, general manager of advertising platforms at Ooyala.

If you’re in the business of selling advertising or otherwise monetizing inventory, you’ve got to follow the money. So it’s hard to argue that the increasingly popular process of header bidding – leveraging a tag in the header of a publisher’s website – hasn’t proven a valuable addition to the digital advertising mix. This yield-enhancing programmatic technique has enabled publishers to offer inventory to multiple ad exchanges simultaneously.

Header bidding’s proponents cite increased transparency and efficiency and a framework that offers publishers greater control and levels the playing field for the demand side. Attractive outcomes, to be sure — especially for display-focused publishers. Companies have even announced plans to bring header bidding to native advertising.

Most notable from my vantage point, however, is the recent wave of announcements from some prominent companies seeking to extend header bidding into the video-ad marketplace. It makes sense, of course, at least on the surface; various sources peg video’s growth trajectory as surpassing other forms of digital advertising.

And yet header bidding, for all its attributes, wasn’t created with video in mind. It’s quite simply an entirely different conversation, for many reasons.

Let’s start with the least important, but most obvious, reason: There is no “header” in a video call – it’s a player.

And then there is channel conflict, a lingering problem in the programmatic video space wherein sellers transacting around the same inventory pool might find themselves working in competition if not duplicating efforts. I’m not suggesting this should stymie programmatic’s general progress, but it is a reality to be acknowledged when plotting ad transactions and relationships. And it is only amplified for publishers selling advertising around long-form premium video content.

Header bidding doesn’t work for premium video because if publishers cannot curate the user experience – or manage the adjacencies of their ads – then they’re putting their user relationships, brand relationships and, ultimately, themselves at risk. Video advertising is by its nature more complex.

For example, some display publishers are trying to adopt out-stream, rather than in-stream, video. The idea here apparently is to develop a near-term revenue boost by taking advantage of video CPMs without shifting overall content strategies toward video.

This is a relatively shortsighted approach when considering the impact on the experience, and may be delaying the inevitable — namely a more radical transition to video content. If publishers are interested in developing a robust video business, they need to think more strategically about evolving their content curation and development and optimize for the user experience.

Header bidding also poses the threat of data leakage. For all the publishers’ success in the NewFronts, buyers have a growing predilection to buy differently – leveraging their powerful and asymmetrical data position. Publishers are balancing their buy-side partners’ embrace of the latest technology and tactics with preserving the value of their inventory – again, a particularly sensitive issue as related to video ad inventory.

One perennial issue with header bidding for display is latency. Viewers already feel strained by buffering and lengthy video load times. IAB-compliant VPAID tags can seriously exacerbate this problem. With header bidding, upward of 10 concurrent transactions happen in real time on each page as it loads, burdening the load speed and ultimately leading to a subpar or worse user experience – again, the lifeblood of monetizing video.

With video, the problem is not solved merely by moving to server-to-server connections or wrappers. Whether on a user’s browser or a remote server, the exchange of all this data requires a lot of heavy lifting from a computing perspective. Ad load – critics might say “clutter” – is what all too often puts the “late” in “latency,” in turn causing users to say “later” to video publishers and brands.

Finally, there is a misperception around header bidding. Some buyers perceive header bidding as access to more premium inventory. Whatever the case for display inventory or out-stream video, for premium in-stream video inventory this just isn’t true. There are no shortcuts here. The intimacy between buyers and seller cannot be usurped by an intermediary, as the publishers have a much more complicated and important role to play, balancing revenue growth with user experience.

As in other industries, automation in video advertising is a positive development – when applied intelligently.

Follow Ooyala (@ooyala) and AdExchanger (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Jeffrey Gilbert

    There are a couple flaws in your arguments here. First, the player for video is the same concept as an ad tag, at markup, iframe or “ad container” so to say it can’t work based on technology is a fundamental misunderstanding of how the mechanics of header bidding technology work. Based on this, it does not come as a surprise that the second most glaring issue with this article is about poor performance and user experience suffering because of header bidding.

    Calls are placed in the head of the document because that’s where there is no DOM manipulation taking place, and are done so with the async attribute which enables browsers to fetch bids for ads from partners in parallel with one another. Yes this has limits (32 in parallel), but this is no better or worse than existing solutions where publishers will drop/shop requests around with different floor prices until they find the best buyer price, much of which is duplicated outbound calling and slow by nature of “wait and see” mechanics. Also, dedicated video requests gain nothing over header bidding implementations because the requests outbound for bids on ad containers/slots can piggyback on existing display calls, making these additional calls close to free.

  2. Aden Forshaw

    I think what Scott is referring to is the prevalence of VPAID wrappers, needed for any pre-bid viewability, anti fraud, player size validation etc – at least until we see VAST 4.0 rollout.

    The real point here is not the tech arguments, which can be solved in time. It’s that any premium publisher should already be selling out their pre-roll inventory via PMPs, and should not need access to lots of Ad networks or SSP’s rtb marketplaces in the way that header-bidding provides.

  3. Aden…perfectly said. The point is much less about whether it can be done, but whether it should be.