A recent Digiday article set out to track down the origin of the term “header bidding.” While the question of who first coined the term remains shrouded in MadTech lore, it’s clear header bidding reached a tipping point two years ago and upended the digital media industry. Now, there is a focus on moving out from publishers’ webpages into AdTech partners’ infrastructures (from browser-based to server-to-server) causing more disruption within the industry. This shift to server-to-server will force adTech companies to work in a more open and transparent manner – or risk losing relevancy. In an industry grappling with proprietary technologies, lack of interoperability, black boxes, and a general unwillingness to play nice with one another — server-to-server header bidding is going to make for some interesting times.
The AdTech industry evolved around publishers’ ad stacks structured in a cascading order of prioritized demand partners, commonly referred to as the ‘waterfall’ set-up. Within the waterfall, Google’s Exchange enjoyed preferred access to inventory for publishers utilizing its DFP ad server. As impressions moved down the waterfall, exchanges further downstream had fewer opportunities to bid on premium inventory as Google and other upstream-prioritized exchanges snatched up impressions. For years, success for many companies was tied to their ability to negotiate a higher placement in, or exclusive access to, publishers’ waterfalls. Within the last two years, header bidding has completely upended this business model. It has shaken companies once considered industry leaders, given rise to new players, and made Google stand up and listen.
With header bidding, publishers offer impressions to multiple demand partners in parallel by placing their tags within the header tag of a webpage before a call is made to the ad server. Publishers use “wrapper tags” to manage multiple header bidding tags on a page, mitigating issues such as latency, and facilitating reporting. This tactic has grown from a way to capture incremental dollars to forming a central part of publishers’ revenue strategies. Header bidding and wrapper tags have allowed publishers to realize increased yield and attain double-digit eCPM lifts. It’s estimated that over 70% of U.S. publishers have a header bidding strategy, with most implementing “browser-based” solutions, wherein much of the operation occurs on users’ individual web browsers, video players, and mobile apps.
The result of all this browser-based AdTech is that content loads slower and places greater demand on devices (i.e., greater usage of mobile data plans and shorter battery lives). This all culminates in publishers sacrificing user experience for higher yield, resulting in a delicate balancing act for publishers looking to maximize revenue while also trying to prevent site visitors from abandoning slow-loading webpages and apps. Additionally, getting competing AdTech players to work well with one another within publisher webpages is challenging.
Browser-based wrappers require one of a publisher’s adTech providers (typically a demand source as well) to manage the header bidding tags of their other demand sources. Having a header bidding tag on a publisher’s page, even when “wrapped” by a competitor’s wrapper, gives adTech vendors a degree of control and visibility into the publisher’s impressions and users. Operations (such as cookie syncing with the publisher’s users and investigating monetization issues) are all simpler when a tech vendor has their code on a publisher’s webpage.
However, even with the prospect of remaining on a publisher’s page within a competitor’s wrapper, AdTech vendors fiercely jockey for position as they each vie to have the publisher utilize their wrapper to manage competitors’ header bidding tags. The jockeying has led to situations unfavorable to publishers, such as some adTech vendors refusing to have their header bidding tags managed within a competitor’s wrapper. As a result, while there are many header bidding tags available to publishers, most proprietary client-side wrapper solutions only offer access to ≈10-15 within their solution. Some publishers have tolerated this situation because many only run 5-10 header bidding partners to limit latency and poor user experiences.
This inability for proprietary header bidding tags to place nice with each other is in part what has driven the broad adoption of Prebid.js. Prebid.js is an open source wrapper solution, supported by a community of developers, and it has access to ~60 demand partners. The open source nature has alleviated some concerns of tech providers and publishers regarding fairness and transparency. As such, AdTech vendors have been more open to allowing their header bidding tags to be wrapped in Prebid. This in turn has attracted many publishers to implement Prebid as their wrapper solution, which has in turn incentivized ad tech providers to continue to support and maintain their header bidding tags within Prebid.
While open-source has been helpful in driving cooperation, the current push to server-to-server header bidding will force further –and much needed– cooperation among vendors. Many of the operations which existed on a publisher’s webpage move within the tech vendor’s infrastructure in a server-to-server environment. This shift removes the visibility and advantages of remaining on the page from competing tech vendors. The mechanics of server-to-server will require adTech vendors to operate in a more open and transparent manner for publishers to fully leverage its benefits. And just as browser-based header bidding unseated leaders, gave rise to new players, and became an ‘oh sh*t’ moment for Google — server-to-server header bidding is set to become even more disruptive.