Media Post, by Tobi Elkin, March 25th
Facebook’s announcement this week that it would officially enter the header bidding space wasn’t all that surprising, given the fact that the platform has been testing the system for months. However, what’s a little more surprising is that it’s turning to partners including AppNexus, Amazon Publishers Services, Index Exchange, and Sonobi to help it build the technology.
What this means is that publishers that work with one of Facebook’s pre-approved partners can gain access to the platform’s Audience Network via existing integrations. Facebook is relying on partners to ensure that Auidence Network demand arrives seamlessly onto a publisher’s site.
Facebook, which also doesn’t currently offer server-to-server header bidding, is looking to its tech partners on header bidding and in so doing, may find ways to nurture stronger relationships with publishers. Facebook’s play drew mixed reactions from ad-tech executives.
“This move from Facebook may finally reveal more of a power shift back to the content creators, who provide the real juice for their platform,” Paul Lentz, SVP, publisher & business operations at ShareThis, told Real-Time Daily via email. “Facebook is now bringing their media buyers directly to the publishers via header bidding and while limited to mobile, it opens up more opportunities for publishers to monetize. This is an encouraging move by the platform to deliver more cash value to publishers on their mobile pages,” Lentz said.
Lentz said that Facebook typically demands a much higher CPM for mobile impressions. “In some cases, publishers will earn a higher CPM on a revenue share from Facebook than they could on their own via programmatic media.”
Others weren’t convinced. Dvir Doron, CMO of video ad-tech firm Cedato, told Real-Time Daily that “Facebook’s ad-tech guru David Jakubowski’s assertion that ad space goes to those willing to pay, is short-sighted since this isn’t the way programmatic video works.” Doron said the way that video is transacted, “any technology that optimizes based on price will automatically fail to address important validation and auctioning issues, as well as having a relatively low yield.”