Today's column is written by Emry Downinghall, general manager at StudyBreak Media.
Most people who are plugged into header bidding agreed that it was only a matter of time before Google made its play, but I don’t think anyone expected it to happen so quickly.
Immediately after Google announced in mid-April that it would open Dynamic Allocation to outside sources, publications began writing death-knell posts, prophesizing the end of header bidding.
Though I’m entirely unwilling to co-sign that sentiment, Google’s emergence in header bidding is significant and certainly disruptive to the current landscape. However, I believe it will ultimately lead to positive developments for client-side integrations, for the tech associated with them and for publishers willing to entertain a hybrid approach to header-bidding integrations.
To understand the impact of this move, it’s important to focus on what was actually announced by Google and what we already know. In its official statement, Google said that it has “begun testing exchange bidding in Dynamic Allocation.”
That’s a pretty broad statement, and while it certainly solidifies its plans to throw its hat into the ring, it’s safe to assume Google’s bidding product is in pre-alpha to alpha stages. As such, it may be far from a feature that’s openly offered on the Google exchange.
If a general release is still far down the road, why make this announcement now? With all the hype surrounding header bidding and solutions being offered by every major exchange, it was in Google’s best interest to make it known that it too intended to roll out an offering.
This was beneficial in two ways. First, it might deter large publishers that were hesitant to adopt client-side header bidding from leaping into the space. Second, it further establishes a platform to discuss the downsides of latency, the most common side effect of client-side header bidding. Latency is important for all publishers to consider and closely monitor, but it’s especially critical for publishers with a high percentage of direct sales.
One thing that was very clear in Google’s announcement is that it plans to integrate these exchanges via server-side connections. While the server-to-server header integration isn’t a new or unique concept, the ability to integrate these connections without developer support or adding code to the page, with minimal latency and unified reporting, is incredibly appealing and warrants consideration from every publisher. In terms of latency, I predict 200 milliseconds will be the maximum bid response time Google will permit.
So what happens to client-side header integrations? As promising and enabling as I believe Google’s platform could be, I strongly believe client-side integrations will remain relevant and, in some cases, thrive.
Innovative publishers are already working with header-bidding partners to increase their capabilities. Concepts such as dynamic flooring, predictive bidding and partner throttling are advanced, exciting and capable of further improving the lift already provided by header bidding. While exploratory features like these require some development-focused heavy lifting, that’s something many publishers will embrace if it also drives yield.
I also believe that Google touting the very real benefits of server-to-server integrations will necessitate a client-side response, where a similar integration will become the norm for all partners. Of course, this would require some SSP partner cooperation which, to date, has been mostly absent in regard to playing nicely with header wrappers.
Why integrate outside of Google if Google is already offering server-to-server?
Perhaps a publisher could find a more favorable revenue share as a result of a direct integration? What if an exchange doesn’t open up all of its inventory to the Google integration, or it awards last-look auction priority to inventory from AdX or AdWords? Will private marketplace deals with other exchanges be supported? How will the auction mechanics play out?
It also may be in a publisher’s best interest to keep impactful on-page integrations live, in order to more effectively drive auction pressure. Setting up thousands of price priority line items is tedious, and it’s next to impossible without script automation, but it’s still necessary for helping to set more accurate and competitive price floors in AdX.
At this point, Google’s emergence in header bidding comes with more questions than answers. However, if its presence improves acceptable latency standards while motivating exchanges that wish to stay directly integrated with publishers to enhance their offerings, header adoption and effectiveness will only increase, and publishers stand to gain.