OpenX Meta Creates Server-To-Server Alternative To Header Bidding Wrappers

OpenX-MetaTo help publishers manage multiple header bidding partners, OpenX created a server-to-server wrapper dubbed OpenX Meta, which it unveiled Thursday.

The advantage of OpenX Meta is that it moves all the action – long, slow bid requests – off a publisher’s page and onto its server. From there, it can quickly ping all the other partners. As a result, OpenX projects that publishers that implement its solution would see a faster ad call process.

Header bidding code normally resides in the browser because it’s easier to set it up that way. RTB started out conducting auctions in the browser. OpenX decided to pursue a server-side solution because this approach limits speeds in three ways, explained OpenX CRO Jason Fairchild.

One slowdown occurs because the header bidding calls to different exchanges are made according to the user’s network speed, which can be particularly slow on mobile. The header bidding calls are also limited by the processing speed of a user’s phone or computer. Finally, a browser may only have 10 ports, according to Fairchild, creating a bottleneck if there are too many header bidding partners vying for a spot.

But if those calls to different partners happen on OpenX’s server, as they do with the Meta solution, they happen at lightning-fast data center processing and connection speeds.

With Meta, OpenX hopes to turn header bidding from a hack to get around an ad server’s limitations to a more sustainable solution that addresses latency and partner management issues.

But it still has plenty of work to do. OpenX must first finish integrating buyers into Meta. Once those are complete, it can approach publishers to implement.

Among the publishers that have expressed interest in signing on is study guide website Slader, which uses a wrapper solution to manage its seven partners.

“As we continue to layer on header partners, we layer on more and more network requests,” said Scott Kolb, co-founder of Slader. “That takes up more bandwidth and more things can go wrong.”

For Slader to use Meta, Kolb will need to make sure overall performance, including both speed and yield, is the same or better. Kolb also needs to ensure enough of its partners plug into the OpenX Meta solution for it to make sense to use.

OpenX admits some tech partners don’t like being fenced in by another’s wrapper. It’s been particularly tough for publishers to bring Amazon and Criteo into their wrappers, for example.

To win over publishers and their header bidding partners, Open X promises that it won’t do two things that can bias an auction in its favor. It won’t give itself “last look,” or an ability to beat the winning bid, and it will pass all bids into the ad server, allowing a publisher to monitor performance of all partners transparently.

Given the “aggressive adoption” by publishers of header bidding, OpenX projects that 25% of its 150 to 200 header bidding partners will sign on to use Meta this year, said Qasim Saifee, the company’s SVP and GM of monetization.

“Header bidding has been a great means for publishers to level the playing field,” Saifee said. “This doesn’t change that dynamic, it just makes it easier for publishers to do it.”

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. I have been waiting for this to happen. Initially this competition had all been done server-to-server, since it was so efficient (RTB was originally s2s, just ask any of the early players like Invite, Turn, Pubmatic, Admeld, Rubicon). Somewhere along the way a few key buyside companies were able to convince publishers to put tags directly on the page (no surprise Amazon and Criteo are fighting hardest here!), promising better returns than via RTB, claiming it was better for both to cut out the exchange costs. Over time every other buyer wanted to get in on the act, but the inefficiency of doing this all client-side has begun to creep in.

    The question is, at what cost does running a transparent (including fees) exchange make sense for publishers and buyers? Client-side bidding is slow and not especially user-friendly. There is a benefit to doing server-side. What rate would pubs feel is fair? Does it need to be a first price auction too?

  2. Doesn’t this alternative solution lead us back to single server dependent integration leading to minimal transparency. Additionally, how would this be different from DFP-First look from Google?