Google lets the users do the dirty work

Plus the latest from OpenX

After four years, Google makes an abrupt U-turn on its plans to deprecate third-party cookies. We are forced against our will to write a newsletter about this.

Ad

Learn how IRIS.TV and PMG helped Carl’s Jr. use video-level contextual targeting in CTV to drive superior lift for visits, sales, and ROAS. Learn more…

The Vendor Interview: OpenX

Marketecture Vendor interviews are free for a week, then require a subscription. Read more about subscription options at marketecture.tv.

SSP stalwart OpenX has been under the radar lately. We catch up with them to hear about their creation plans and continued focus on CTV.

Podcast: Mario Diez on the state of contextual

What's the state of the art in contextual advertising? How does AI fit in? What happened at Oracle? These and other questions are answered by Pier39 CEO, Mario Diez.

Listen to the pod now:

Google lets the users do the dirty work

By now, surely you've heard about the news that Google is no longer going to be removing third-party cookies unilaterally, but it will instead present end users with some sort of privacy dialog. We had a great conversation about this on a "emergency podcast" featuring myself along with Andrew Casale from Index and Paul Bannister from Raptive. Let's go through what we know and what we don't know, and some smart takes from “friends of the pod.”

Conventional wisdom

By now most pundits have come to more or less it consensus, that this news changes very little. Google is essentially still deprecating cookies just blaming it on the end users instead of taking responsibility itself.

We have zero insight into the nature of the Chrome implementation or interaction with consumers that's relevant to third-party cookies. But we know that Google’s intent is to get rid of these cookies because of their continued battering at the hand of rival apple (see meme below). We also know that Google believes that third-party cookies represents a real privacy issue as they've been saying that for 4+ years.

The regulatory deck gets shuffled

In hindsight, the Sandbox initiative was hugely ambitious. It required not just completely reworking the infrastructure of the large and complex ad tech stack, but also gaining wide adoption from parties with divergent interests.

It turns out that one group most affected by the plan struck back by enlisting regulators. Thus, CMA oversight over the cookie and Sandbox replacements became quite focused on the likely effect on publisher revenue, which, as we've been following, is certainly going to be negative (latest estimates below).

The most recent CMA report (pdf) brought up numerous fundamental issues about the Sandbox as impediments to roll out, and, as we reported at the time, these problems were not new and had been unsolved for the four year lifetime of the project. For example, the need for the Sandbox to be governed by a third-party was something that Google and others had been saying for some time, but which has not made any progress as far as we can tell. If the future of this initiative is dependent on changing quite fundamental things about the initiative that have not been changeable to date that's not a great prospect for actually releasing this at any time in the near future.

The CMA has decided to withhold publication of its upcoming report and instead ask the public for commentary. No one (other than Google, presumably) has seen this draft report, and now no one ever will. There have been rumors swirling that it was going to be quite damaging to the future of the project.

So the billion dollar question is whether Google can get out of it regulatory bind by shifting the control over cookies from their engineers to the end users. This may seem a little ridiculous, but there is a direct comparison to the way Apple rolled out ATT, which did not have any regulatory input. I’ll leave you with this tweet from Shiv:

Impact on advertising

I hate to be wishy-washy, but this announcement is both positive and negative for the overall advertising business. Here are some points of view (all mine):

Nothing changes: We don't know what Google’s approach will be, but we do know that there will be far fewer third-party cookies a year from now. Which means that we should continue the evolution of our stacks to a cookieless future even if it's not really cookieless.

More signal is good: Whether 10% of users or 30% of users still have third-party cookies a year for now, those users will represent an important signal that can be used to train our cookieless stacks. Use cases ranging from retargeting to attribution to media mix modeling will have a data rich stream to use for AI training and validation.

Lazy marketers continue to coast: Those in the trenches have seen how marketers and agencies have failed to date to step up to the challenges of a cookieless future. Numerous surveys have shown that folks are really not prepared for this eventuality. The worst effect of this news may be that it gives a further excuse to over-target the minority of Chrome users with cookies.

Other smart takes

Eric Seufert writes about ($$) the fate of Android’s GAID, which was left outside of this week’s announcement. In 2022, Google said that GAID would be deprecated in the same or similar fashion as the Chrome cookies. To me, this seems pretty different because it is unlikely to have as much opposition from publishers and there's a clear precedent from Apple.

It’s important to recognize that the GAID will be eliminated either way. But the text of an opt-in prompt likely has a meaningful effect on opt-in rates, and I don’t think that Google is as incentivized to influence low opt-in rates with Android as it is in Chrome.

—Eric Seufert, MobileDevMemo

Another shoe to drop could be Google’s pre-existing consent agreements with the regulators which were largely tied to the framework around deprecation. For example Google has agreed not to use certain media from within Chrome to benefits advertising business. One wonders if these may also be to renegotiation.

Appendix: The latest Sandbox data

Less noticed within Google’s announcement was their data on Sandbox effectiveness (in bold). We can add their data to our ongoing tracking:

Data source

Decline in CPMs

Criteo

60% with Sandbox

78% overall

Raptive

30%, no Sandbox

Index Exchange

33% with Sandbox

Google

20% with Sandbox

Safari, baseline

60%

Reading list

  • Reddit not only licensed its data to Google for AI, they are restricting search results on other search engines. Both a competitive nightmare and an illustration of how AI and search are colliding. (link)

  • AdWeek reports that Pluto TV is being accused of excessive RTB auctions and auction manipulation by buyers. (link)

  • Peter Naylor out at Netflix. Jeremi Gormand was ads boss in 2023, replaced by Netflix “insider” Amy Reinhard. (link)

  • Kerel Cooper exits Group Black to join Gumgum as CMO. (link)

  • Epic article this week in BI by Lucia and Lara on the rise and fall of Group Black. (link)

  • Oracle gets big fine for privacy violations (link)

  • Jon Suarez-Davis had a good post on this on LinkedIn (link)

  • Outbrain rumored to be an acquirer of Teads (link)

Reply

or to participate.