Today’s column is written by Danny Khatib, co-founder and CEO at Granite Media.
Something important just happened in the header bidding world that you may have missed.
AppNexus and Rubicon recently announced the formation of Prebid.org, an independent organization to manage Prebid.js and other open-source projects related to header bidding. Some regarded the announcement as an “old-news, nothing-to-see-here” story, a procedural update rather than a strategic shift, mainly because AppNexus had already made the Prebid.js project feel like a community-driven open-source effort.
But I see this as a graduation moment for Prebid, and we should celebrate the milestone as the new entity heads off into the real world on its own.
Why Should We Care About Prebid?
Open source efforts like Prebid provide the opportunity to commoditize underlying technology and reallocate capital toward more complex services higher up the stack, which the supply side sorely needs. We don’t want publisher and supply-side platform (SSP) developer teams devising unique solutions to the same demand piping problem. We want them focused on new modules for analytics, auction mechanics, data integration, native template rendering and other forms of extensible functionality.
There is also strategic power behind a boots-on-the-ground coalition focused on software development and implementation rather than political lobbying. When coalesced around a single effort like Prebid, we can work separately but in concert to seek additional requirements from tech providers, and our collective voice is more likely to be heard. It’s not enough to have an OpenRTB spec; we need an easy implementation mechanism and a common vocabulary to rally behind.
Many SSPs and ad tech vendors are fighting battles on several fronts and stretching for resources. A shared initiative like Prebid could help actors collectively stretch their capex dollars further in pursuit of competitive advantage.
Prebid and Prebid hybrids have gained impressive penetration among top publishers. Publishers with fully custom, in-house wrappers are likely the leading tech-savvy publishers that built their own solutions before third-party alternatives like Prebid were any good. It may not yet be worth it for them to switch over, but eventually it will be as maintenance costs increase and compelling Prebid modules emerge. Any publisher building a header-bidding stack from scratch can implement Prebid within days. The adapters are a lot less ugly now, so it wouldn’t make sense to write them all from scratch.
The Leading Wrapper Integration Options
People in the trenches will say it’s a four-horse race from a wrapper infrastructure perspective, ranging in openness from Prebid, which is fully open, to Index Exchange, Amazon and Google, which is increasingly closed.
Prebid will come in two flavors: publisher-implemented and partner-implemented.
Pure Prebid is favored by publishers that want to do the work themselves, understand the internals, tinker and favor fully open architectures for strategic reasons. If you want complete transparency and control and are willing to do the work, this is your best option.
PreBid hybrids from AppNexus, Rubicon, Pubmatic and others present a compelling variant for those wanting a more full-service multi-SSP offering, the benefits of understanding and modifying internals and Prebid’s code of conduct, which ensures open design and fair auction dynamics. Many SSPs will evolve their wrappers to become PreBid hybrids under the hood for these reasons. If you want to support openness and transparency while letting a partner handle all service aspects, such as yield management and reporting, this could be a fine option.
Index Exchange remains a solid proprietary option for those wanting a fully managed offering and a partner that primarily serves publishers. Index Exchange’s multi-SSP strategy has gained traction because of its reputation for transparency. It also benefits from a clean financial structure, allowing it to focus on the long term. It will need to keep up with rising capex requirements and could buddy up to Prebid or devise its own open tech strategy. For now, as a technically closed but culturally open option, Index is a stable candidate that fits squarely in between Prebid and Amazon.
Which Duopoly Are We Talking About?
Typically, people talk about Google and Facebook as the duopoly in the ad space. While that’s true for advertiser spend, the emerging duopoly in the supply-side technology ecosystem is actually Google and Amazon.
Amazon once looked like Facebook when A9 was just a large buyer going direct to pubs, cutting out the middleman. Because it was early in this game, A9 was integrated directly onto a lot of pages. Amazon’s demand business – Amazon Audience Platform – has quietly become a DSP powerhouse. In classic Amazonian fashion, the company has begun closing the platform loop with a multi-SSP solution called Transparent Ad Marketplace (TAM), which integrates demand from other SSPs alongside Amazon’s internal demand.
It has very quickly gained serious traction, driven by integration pipes already in place and Amazon’s brand power. Everyone assumes Amazon will be successful, which is a bit of a self-fulfilling prophesy. Amazon also offers a low-cost solution by charging almost nothing for facilitating third-party demand – again, classic Amazon. It’s also easy for publishers to turn on incremental demand, as Amazon handles almost everything server-side. Prebid is working on a server-side solution, too, but Amazon is ahead in this game.
Amazon’s open demand solution is coupled with a closed technology approach. Publishers aren’t typically allowed to integrate Amazon into Prebid or other client wrappers. TAM feels more like a DoubleClick competitor than the sister division of Amazon Web Services, which has a radically open platform and culture. Publishers attracted to Amazon want ease of implementation, quick demand acceleration and the stability of a massive organization. They are less concerned about transparency, control or the strategic implications of empowering another Goliath.
On the fully closed end of the spectrum is Google EBDA, a pseudo-wrapper option since it just uses DFP server-side. It’s not technically a wrapper but it tries to serve the same purpose of aggregating and piping demand, so it’s worth including here.
Between AdWords, DBM, Ad Exchange and DFP, Google has several competing incentives at play. Google is a strong partner, but it isn’t a champion for publisher transparency and control. Publishers attracted to EBDA want as little operational friction as possible, regardless of whether it squeezes out the most absolute long-term yield. Some are quietly testing EBDA but are very worried about giving Google back the strategic high ground it recently lost due to header bidding.
Regardless of which option publishers choose, let’s all take a moment to celebrate the official introduction of open-source thinking and methodology into the supply-side ecosystem. Open-source solutions have catalyzed innovation in many other industries. Let’s hope that Prebid.org can do the same.