In The Name Of Efficiency, Buy- And Sell-Side Companies Create OpenRTB Standard For Online Display Buying

By
  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

OpenRTBIn an Ad Age article on Friday, a group of buy- and sell-side companies announced that they had come together to form a new, open standard for online buying they call "OpenRTB." Read the article. AdMeld, PubMatic, Rubicon Project, MediaMath, Turn and DataXu are the initial group members. Ad Age's Edmund Lee explains, "The emergence of a single standard around how both DSPs and RTBs communicate in the ad-buying process underscores the fact that more and more publishers and ad buyers are transacting on these platforms." Visit OpenRTB.info.

Bill Simmons, CTO of DataXu, discussed the new standard and its meaning.

AdExchanger.com: Why form OpenRTB - what problem is this solving? What are the benefits of banding together?

BS: We led the creation of OpenRTB to remove barriers to the growth of programmatic buying and selling of interactive advertising.  As a starting point, we are removing some of the common technical roadblocks that slow down the growth of the real-time bidding market for demand-side platforms (DSPs) and sell-side platforms (SSPs.)  Specifically, we are agreeing on common standards and techniques for synchronizing advertiser and publisher preferences.   In addition, we have agreed to some common taxonomies for labeling of creatives and agreed on standards for filtering out publisher prohibited ad serving and tracking vendors.

The specific benefits of OpenRTB are:  1) Higher CPMs for publishers, since bidders will have accurate pre-filters, 2) more accurate and cost efficient bidding for buyers, 3) better enforcement of publisher and advertiser restrictions, due to a common way of specifying them, and 4) easier integration of DSPs with SSPs, since one you have set up OpenRTB once, the cost of setting up a new connection is trivial.

Overall, this is an "everyone wins" situation that is good for advertisers and good for publishers, good for DSPs, and good for SSPs.

More specific benefits can be found at our project website at http://openrtb.info.

Can you give a specific use case as to how OpenRTB will come into play?

OpenRTB V1.0 is focused on synchronizing publisher preferences with a bidder’s filters.  For example, let’s say the WSJ.com doesn’t want alcohol, tobacco, competing news ads, and a direct-sales account (let’s say Toyota) appearing on their site through the RTB exchanges.   With confidence, they can specify these requirements in a unified way using OpenRTB standards.   The same applies to advertiser who had requirements for the types of sites where they may not want their ads to appear. For example, McDonalds may not want its ads in a health & fitness or vegetarian  site or near articles on child obesity and high cholesterol.

The exchange (an SSP) will set up a simple OpenRTB web service so that all potential bidders will be able to download the publisher blocklist.  In this way, all bidders will know ahead of time to not bid on impressions on behalf of any advertisers who do not satisfy those restrictions.

WSJ.com can now put its inventory into the RTB marketplace with confidence, since they know their restrictions will be enforced with precision. With more bidders sending correctly pre-filtered bids, they will see more valid bids per impression, and their CPMs will rise.

The OpenRTB working group is supplying source code and reference implementation to make it as easy as possible for new players to use the system.

What about Google's DoubleClick Ad Exchange or Yahoo!'s Right Media Exchange? Will they be included? Is it intended as competitive to their exchange efforts?

OpenRTB’s mission is to create open standards. We kept the initial group small to ensure we were able to get version 1.0 out quickly.   Now that version 1.0 is out, we are fully expecting companies large and small to join and contribute their ideas. There will be exponential benefits as more companies join.

Since news travels quickly, there have already been some initial discussions with additional partners but, since we have kept the V1.0 specification under wraps until it was complete, they haven’t had a chance to join yet.  We hope that the effort continues to build momentum so that we have an open source set of buying standards used by many but not controlled or dominated by any one party.

So, to be clear, this is isn't about a private exchange or closed group of buyers and sellers banding together?

No, it’s just the opposite. We like to analogize to the formation of the New York Stock Exchange, which began under a buttonwood tree outside 68 Wall Street in New York City on May 17, 1792. On that day, 24 competing stockbrokers signed the Buttonwood Agreement creating the New York Stock Exchange, a governing body of brokers to establish guidelines for the purchase and sale of public stock.

Up until the signing of the Buttonwood Agreement, there had been no universal standards for the buying and selling of stock. The Buttonwood Agreement brought together rivals in a marketplace who understood they’d all have a better chance to prosper if there were common standards governing their transactions.  Similarly, in today’s digital advertising world,  OpenRTB is about eliminating the friction in the ecosystem in order to pave the way to massive growth of programmatic buying and selling of digital advertising.

Will you add other DSPs, exchanges or supply-side sources as OpenRTB partners? What will be the criteria?

If other exchanges, DSPs or supply-side companies would like to join OpenRTB, they should contact us at openrtb@dataxu.com.  The main criteria are a willingness to commit resources to implement OpenRTB into their RTB technology stack, and a commitment to contribute resources to the further development of the OpenRTB source code and standards.

Is OpenRTB rolled out among the six partners today? Across how much buying/selling? Any data?

There are some simple prototypes of this system in play today.  Commercial roll out of the official version 1.0 of OpenRTB is scheduled for the end of January.  The six partners have pledged to implement OpenRTB as part of their technology stack.  The goal is for 100% of our traffic to be pre-synchronized using OpenRTB.

Is a universal cookie pool a part of this in order to standardize the buying?

We are not considering creating a common cookie pool.  OpenRTB was intentionally designed as a peer-to-peer system to avoid the need for global synchronization. This allows DSPs and SSPs to maintain their competitive differentiation, while standardizing on the data synchronization tasks we all have in common.

By John Ebbert

  • Facebook
  • Google Plus
  • Twitter
  • LinkedIn

Email This Post Email This Post

By on at

5 Responses to “In The Name Of Efficiency, Buy- And Sell-Side Companies Create OpenRTB Standard For Online Display Buying”


  1. Jack Marks says:

    Seems strange.. Why would the DSPs feel the need to band together?? Seems like the elephant in the room here is Invite/Google..

  2. Sara Livingston says:

    Does anyone know what the approximate % of total impressions bought through DSPs will be covered in the first phase of OpenRTB?

    I know it's difficult to determine given no one really know what % of total impressions are bought on an RTB basis. However, given that OpenRTB only includes a set inventory source(s) and demand is limited to 3 DSPs, it should be easier to come to a ballpark %.

    (also what's the acronym for this, ORTB?)

  3. Ryan Maynard says:

    I still love how real time bidders assume higher CPMs are the end all be all for publishers. If publisher CPMs go up 4x but the fill rate is only 10% because all of the RTB buyers are cherry picking the very best impressions and defaulting/passing back the rest, how does that ensure higher overall revenue for the publishers? 10mil impressions monetized at $0.50CPM ($5000) is still better than 1mil impressions monetized at $2.00CPM ($2000). Show me a RTB solution that can ensure 100% fill rate AND higher CPMs for pubs and I'll get excited

Leave a Reply