Not a slow news week

Google delays cookies, TikTok banned, declines in open web

There was a lot going on last week. We’re doing our best to keep on top.

Vendor interview: Innovid

Following up on their recent “Harmony” announcement, Ari sat down with Chief Commercial Officer Dave Helmreich to review all the company’s offerings and to hear how the acquisition of TVSquared is evolving.

Watch the interview with Ari, free for a week before going behind the paywall.

Reminder: Vendor Interviews are always free for 1 week, then are subscribers-only. Subscribe for only $39/month.

Podcast: WPP’s $300 million AI investment

You might be surprised to learn that advertising giant WPP even has a CTO. Well would you believe they are spending more than $300 million annually on AI?

Stephan Pretorious puts all that doubt to rest with an impressive discussion of the holding company’s substantial investments in AI and tech.

Listen to the pod now:

This was not a slow news week


Lots going on, let’s supply the hot takes around cookies, TikTok, and Google earnings.

Apparently Paul Bannister has psychic powers since he posted this Tweet:

And then literally four hours later, Google posted their statement that the deprecation of third-party cookies in Chrome would be delayed from Q4 2024 through sometime in the first half of 2025.

AdExchanger was in excellent form with two good contributions:

The next day we found out more when the UK’s CMA published their latest interim report (pdf) on the process. The document is a long, interesting read, but I can’t help but wonder about the tone of politeness, as Americans often can’t tell when British people hate us.

In any case, the objections the CMA have seem really fundamental to the overall enterprise. Just on page 6 they bring up three absolutely foundational issues of concern:

  1. That the Sandbox will benefit GAM;

  2. That the continued governance of the Sandbox by Google gives them too much power;

  3. That Google will still yield enormous amounts of O&O data, giving them an advantage.

No shit, Sherlock Holmes. Anyone on Ad Tech Twitter could have rattled these concerns off three years ago when this whole idea was floated. Maybe I’m being dramatic, but it seems pretty late in the whole process to fix these massive problems.

At some point in this process something breaks. Whether this means cookie deprecation and Sandbox timing get decoupled, or the CMA gives in on some issues, or something else, we’ll be watching.

Meanwhile. there has been a definite change in posture around cookie deprecation in 2024, and I believe most industry participants are investing and getting ready for the inevitable. Having another two quarters to do so seems like a blessing, as everything takes time.

TikTok bill signed

The bill gives them 9 months to sell. The company has indicated that they will a) sue; b) if unsuccessful in their suit, shut down rather than allow an American company to own the algorithm. That’s a pretty curious position — are they maybe motivated by something other than capitalistic profits? Hmm.

It will be interesting to see how big an issue this turns out to be in the upcoming Presidential election. Biden isn’t exactly doing well with younger voters, and the TikTok issue has already gotten conflated with the Palestinian issue, resulting in an unpleasant set of facts for the campaign.

Regarding the outcome of a ban, Semafor posted a great chart showing what happened after a ban in India, where Insta took all the share:

Google earnings and the decline of the Network business

Have you noticed that the “Meta” name change seems to have stuck, but no one calls Google “Alphabet”?

This newsletter isn’t the place to go for stock tips or analysis, but the story coming out of Google earnings is pretty interesting. We’re seeing a sustained reduction in the size of their “network” business, which is the combination of AdX, the former DoubleClick products, DV360, and GDN.

Source: @tkawaja

What makes this even more interesting is that it is coinciding with the growth in the YouTube business. Here’s a chart showing the trend, with the lines crossing two quarters ago (from Marketecture contributor Eric Seufert):

There are a couple of schools of thought around these trends:

School 1: This is the point of pMax. Whether you think it is nefarious or just a natural outcome of the machine learning, there is a widespread belief that Google’s automated buying algos naturally favor O&O over network, with the net effect being that more of the ad dollars in Google Ads get sent to YouTube.

This probably could be the subject of its own dedicated newsletter, but there are some reasons why this would be happening.

First, the algo optimizes for Google’s margin, and this is generally going to be higher on YouTube than on the open web since the TAC will be lower. Google still has TAC on YouTube, but only on some of the content and I would have to imagine it would be much lower than through RTB.

Second, Google has more identity and behavioral data on YT than on the open web, and this will be more so the case after cookies go away. Side note: I believe Google has said they won’t use their “Sign in with Google” data to target ads post-cookie, but I would have to assume they do, and will continue to use that data on their O&O properties.

Third, there’s less price competition on YouTube since it isn’t available to other bidders. So when the value of the impression is very high (retargeting) there isn’t as much bidding pressure and margin goes up.

School 2: The open web is in secular decline. I don’t think anyone is really jazzed about the banner ad business, even while we still have cookies. In contrast to Google’s network business, though, public ad tech companies like TTD, Magnite, and Pubmatic continue to grow top line. So how to reconcile?

One explanation is that Google’s Network business is largely missing the boat on the CTV market. Because of restrictive policies on privacy (cough, IP addresses, cough), and a lack of video inventory in AdSense, all the action for video on Google is on YouTube. In contrast, the independent ad tech companies are aggressively growing CTV at the expense of banner.

School 3: The Mohan effect. Google Network reliably grew revenues while Neal Mohan was in charge, then he switches jobs to become the CEO of YouTube and guess what happens? Coincidence? I think not.

Reading list


or to participate.