Fledge, part of Google’s Privacy Sandbox suite of solutions intended to replace third-party cookies, is moving from a theoretical ad-tech framework to reality, according to new findings from demand-side platform RTB House.
But encouraging all members of the media supply chain to test, especially in tough economic conditions, is limiting its adoption.
Fledge, or first locally executed decision over groups experiment, aims to deliver privacy-compliant, targeted ads via on-device auctions and is predicated on buyers, sellers and vendors adding users to interest groups based on browsing activity or behavior.
RTB House displayed nearly 120 million impressions via Fledge auctions in January 2023, an 18-fold increase from the 6.5 million served in August 2022 (though still a tiny fraction of RTB House’s monthly average of 30 billion impressions). RTB House can use Fledge to serve more users with ads in part because Google is expanding its testing population on Chrome.
Industry consensus is growing around Fledge being a viable, privacy-safe alternative to transacting digital advertising via third-party cookies. Fledge live testing has more than tripled since October, according to a Google spokesperson, and 46% of RTB House’s performance advertisers are bidding on Fledge auctions. But the tough economic climate hampers experimental testing and any desire to spend on tech integrations.
Fledge is especially lacking in support from supply-side platforms, which let publishers sell their inventory programmatically, and are especially challenged. While several SSPs are testing Fledge internally, RTB House has found no bid requests coming from SSPs, with the exception of Google’s SSP, Google Ad Manager.
“I hope that other SSPs will do something about this,” said Lukasz Wlodarczyk, vp of programmatic ecosystem growth & innovation at RTB House. “[Otherwise], publisher interest will not be represented.”
In cases when RTB House wants to test Fledge auctions with other SSPs, the company has developed a workaround: When an advertiser wins the auction via RTB House, the DSP runs a second auction via Fledge if the user it’s bidding on is part of the Fledge testing population. RTB House declined to share performance data for Fledge auctions, partly because the firm is not yet able to compare Fledge apples-to-apples to cookies given the lack of available data.
The state of the economy deprioritizes any effort to test at scale because every dollar counts.
Anonymous SSP exec
Otherwise, Poland-based RTB House is also testing by directly integrating with publishers, including European titles such as Ringier Axel Springer, Wirtualna Polska and Ströer Media. U.S.-based ad network CafeMedia, which had already been testing Fledge internally, plans to go live with the protocol in the bidstream in the coming weeks, said its chief strategy officer Paul Bannister.
Two SSP sources told Adweek that while they see the value of testing Fledge, it is not an immediate priority given Google’s changing deadline.
“It’s been difficult to justify investing a bunch of engineering dollars in a moving target,” said Garrett McGrath, SVP of product management at SSP Magnite.
More technically laborious than other alt IDs
With Fledge, buyers don’t bid on individual users but via an interest group where users are members. When a user visits a site, all the buyers who own interest groups that contain that user are invited to bid, and the whole auction takes place in a browser, not an ad server.
Fledge constitutes a totally new system for auctions, and setting it up is more technically laborious than swapping out the cookie for another identifier, said McGrath.
“I wouldn’t conflate fledge with cookie alternatives,” McGrath said. “UID2 or any number of cookie alternatives are not programmatic transaction alternatives. Fledge fundamentally rewires the transaction.”
Sources said that the technical lift was not necessarily greater for SSPs than DSPs, and the fact that more DSPs, like RTB House and Criteo, have tested Fledge is mostly circumstantial.
Missing the boat
But when ad-tech firms are laying off employees and SSPs are closing, some companies may decide not to waste impressions on techniques that are still experimental, said an SSP source who didn’t have the authorization to talk to the press.
“The state of the economy deprioritizes any effort to test at scale because every dollar counts,” the source said.
Others on the sell side are more eager to test. Some SSPs and publisher entities are in the process of setting up Fledge integrations, said Bannister, calling the current moment “the end of the beginning.”
“We’re at the point where this is getting serious, and if you don’t get moving now, you could miss the boat,” Bannister said.
A Google spokesperson said DSPs are ramping up Fledge testing while SSP interest in testing is growing.
Still, many practitioners don’t see the value of testing any one solution, perhaps choosing one alternative at random but without much urgency, the SSP source said.
“It’s been really hard to understand why certain companies value certain opportunities more than others,” the SSP source said, referring to different cookie alternatives. “There’s upside for both the buy and the sell side if Fledge can be figured out. In the interim, it’s a little bit like herding cats.”
