“The Sell Sider” is a column written for the sell side of the digital media community.
Today’s column is written by Paul Bannister, co-founder and executive vice president at CafeMedia.
Just as buyers are now grasping how header bidding impacts their programmatic buying approach, publishers have already moved on to the next big thing: server-side header bidding.
And just as buyers who fully embraced header bidding got a leg up on the market, the same opportunity exists for buyers who are the first to navigate how the new server-side header bidding technology changes market dynamics. A two-tier supply pool will emerge – inventory will be premium in one pool and less valuable in the other. Buyers need to position themselves now to maintain access to high-value inventory.
While server-side header bidding is still in its infancy, adoption will increase dramatically over the next six to nine months and have a major impact on the supply ecosystem.
Yield is often mentioned as the main benefit of header bidding, but this is rarely the case for premium publishers. For these publishers, header bidding provides access and control that can be leveraged in advertiser partnerships. By getting insight into the value of every impression, publishers can prioritize impressions from advertisers with whom they have a direct programmatic relationship.
Publishers can also individually manage private deals in the ad server, which cannot be done through a standard tag-based programmatic connection. And due to the proliferation of header bidders, each publisher has chosen preferred exchange partners that get better access, better management and more data – so advertisers who buy a given publisher through the right exchange get the most value.
Standard header bidding runs code on the web page that collects buyers' bids for a given impression before the ad server is called. The downside of header bidding is that the presence of code on page can increase page latency, harming user experience. As most publishers have multiple header bidders, the issue has been magnified.
Server-side header bidding addresses these challenges by moving many header bidders to the server so the user's browser does not need to execute the code. This means a header bidding partner on a page calls the other exchanges, collects their bids and passes them back to the publisher with its own. This doesn't move all code off of the page, but over the next year, many publishers will shift from having five to 10 header bidders on the page to more like two to four.
Exchanges that are moved to the server side will have to cope with certain limitations that will further widen the gap between the preferred exchanges and the secondary exchanges for each publisher. The exchanges on the server side lose direct access to the user – and therefore some cookie data is lost – and must contend with shorter auction windows. Since publishers enforce an auction timeout in the page, server-side exchanges will inevitably have the least time to respond.
As most publishers have already chosen a small set of preferred exchange partners, server-side header bidding will create an even greater dichotomy between preferred and secondary exchange partners. The exchanges on the page will have the best cookie data, longest auction windows, best access to publisher data and the most attention and proactive management from publishers. The other exchanges will fill a role for publishers as remnant backfill. Buyers cognizant of the widening gulf between preferred exchanges and the rest can use it to their benefit.
This two-tier supply split will become increasingly evident over the next year. Publishers will push their largest clients through the preferred channels and allocate more and more of their best inventory to those buyers. The inventory remaining in the second-tier pool will become less and less valuable to the remaining buyers. Getting ahead of this trend is critical for buyers who want to retain access to high-value inventory.
Buyers who value access, priority and quality will move away from open market buying, where there is no information on what exchange is preferred, to private marketplaces, where they are more likely to get more value on their investment. Private buyers will likely prioritize publishers’ preferred exchanges rather than those preferred by their DSPs as this will also maximize value for them.
Buyers who do choose to remain heavily invested in open-market buying should understand from exchanges what percent of the time they are an “on page” header bidder versus a server-integrated partner. Biasing spend toward on-page bidders at a macro level will maximize a buyer’s chance of getting the better access point, but they will rarely know the details.
Server-side header bidding is radically accelerating existing trends in the supply landscape. Buyers who understand and react to these changes earlier can get a competitive advantage and maximize their long-term access to the best inventory. This new technology is going to have large implications for the industry overall, and buyers need to adjust their strategies accordingly.