Media Briefing: Publishers rail against opening up their first-party data in a post-cookie world

Data protection

The key hits:

• Publishers are wary of cookie-replacing identifiers replicating the third-party cookie’s ills.
• Email-based IDs like Prebid’s Unified ID 2.0 do not provide publishers as much control over their data as they would like.
• Akamai may offer a solution that suits publishers.

When it comes to identity technology, publishers feel like they were fooled before by the third-party cookie. They are determined not to be beguiled by its replacements that depend on publishers opening up their first-party data.

“We don’t want to replace cookies with hashed IDs using email…. We should not be replicating the current system,” said one publishing executive during a closed-door session at the Digiday Publishing Summit that was held last week in Vail, Colorado. 

This executive was far from alone in their wariness of the post-cookie landscape for publishers resembling the cookie era — to the point that one publishing executive half-jokingly wondered whether the assembled publishers risked violating any laws if the conversation turned toward the group mounting a united front against ad tech intermediaries seeking publishers’ first-party data.

“Everyone feels like we already gave up the data once to everybody else. The question is, how do we control it this time around a little bit smarter, not chase the nickel and actually get the dollar?” said a second publishing executive in attendance.

The publishing executives’ concerns boil down to their recognition that their audiences are their oil and identity tech often permits outside companies to siphon that black gold. In other words, enabling ad tech companies to identify and track publishers’ audiences around the wider web — as they have done with the third-party cookie and stand to do with its replacements — risks diluting the publishers’ value in dealing with advertisers by commodifying their audiences.

“If we go down this path of publishers putting their identity into the auction and it gets shared and duplicated the same way that third-party cookies drove [demand-side platforms’] growth, you’ll never get it back and you will lose the last thing to answer the question, ‘why shouldn’t I just buy you in the open auction?’” said a third publishing executive in attendance.

A fourth publishing executive in attendance said the data management platform they use is already aggregating all of its publisher clients to put together “competing marketplace deals.” This executive’s recourse has been “doubling down on the hashed email. We can control hashed email; we can then sync with anyone and try to expose that data when we choose to,” they said.

However, that’s not entirely true. Publishers can only control how they expose their hashed email data to a certain extent. Unified ID 2.0 is among the most well-known identity tech built on hashed emails. But UID 2.0 doesn’t provide publishers with as much control over who their data is exposed to as publishers would like.

“It’s all platform-based. I can enable it for the entire Trade Desk and for [Google’s DSP] DV360. But I want to get down to enabling it by the seat level: I know I have a partnership with this particular agency or group that we have contracts with, we know what they’re doing with it. But the technology just isn’t there yet,” said a fifth publishing executive in attendance.

“Right now it’s a one-to-many [relationship when providing access] and we lose track and credibility and transparency in the current environment — I would like to not do any of that,” said another publishing executive.

So what are publishers to do? Some publishing executives in attendance struggled to see a cookieless future that does not resemble the cookie era with respect to publishers’ data becoming commodified. But others saw hope in one identity tech option that turned several publishers’ heads during a working group session on “Life After the Cookie.”

Akamai has a solution “that actually seems like a really good idea because no data is actually shared. It’s all marketer-centric; they have to come with data,” said a publishing executive during the working group. As this executive explained, Akamai would use its role as a content delivery network that serves content across the web to effectively function as its own clean room so that an advertiser would provide an audience segment of people it wants to target with ads and Akamai would handle the targeting without a publisher needing to share any email addresses.

“That is something I’m rooting for because that allows marketers to target who they want to target, we don’t have to give them data and, in theory, that just seems like the best way to do it,” said the executive.

But there’s a but. You may have noticed two words in that quote: “in theory.” Therein lies the rub, as another publishing executive in the working group identified: “We like the Akamai approach as well. But there’s no money there yet.” — Tim Peterson

What we’ve heard

“We were working with one vendor where we had to show them that they were out of compliance.”

— Publishing executive

3 questions with The New York Times’ new team working to personalize the homepage

The New York Times is testing a transition from a more manual approach to programming the homepage of its site and app to one that will incorporate algorithms to personalize the presentation of stories.

Coinciding with that transition, The New York Times announced last week the formation of the “experiments and personalization” team. Part of the Times’ “home” team, the new group will work with editorial desks and product teams to test algorithms designed to get readers to engage with more content and continue returning to the homepage, such as targeting readers based on their location or reading history.

The team is led by Derrick Ho, deputy editor for personalization. Digiday spoke to Ho and associate managing editor Karron Skog, who oversees the home team, to find out why the Times is putting more resources into experimenting with personalizing the home page for subscribers.

“We are using data to help us make decisions, in an editorially-informed way,” Skog said. — Sara Guaglione

The interview has been edited and condensed. 

What are the specific responsibilities and priorities of this new team?

Skog: We publish something like 200 URLs a day that my team is sifting through, reading, evaluating and programming onto this very finite surface of our home screen and our app. No reader can get through 200 pieces a day. We are trying to use some of this work to really put the right things in front of the right readers at the right times. The top of our home screen and our news priorities for the day are the same for everyone and will be the same for everyone. Our readers value the Times’ editorial judgment. So that’s really quite sacrosanct. I think that there are ways, though, that we can help readers find this journalism. We’ve got subscribers who come to us once a week, and we have people who are coming to us five to 10 times a day. How do we program for those two very different types of readers? The experimentation is trying to help do what we can’t really do in manual programming. We’re constantly putting things on the home screen and taking things off the home screen and moving things around. Making sure that readers see the things that we think are important on any given day, no matter when they visit us, is a big piece of this work. For me, that’s something that we wished we could do for a really long time. And the fact that we have a team now devoted to really trying to make that happen is really exciting.

Ho: We’re introducing a different mode of programming here. We are still in the very early stages, but we [have been] given the room and space to experiment and try new concepts and think about new ways we can amplify our editorial judgment and also meet reader needs. There’s also room and space for us to work with product. We’re still very much in the phase where we are trying to build the tools and refine the tools. We are researching and doing a lot of user research.

Homepages were all the rage a few years ago. Why the focus on this now?

Skog: I have been working on the home screen at The New York Times for a very long time. And we’ve gone through many phases of how we program it. Eight years ago, people came to us once a day. So we needed to leave things on the home screen for a very long time. And then a couple years later, people were coming many, many times a day, and we felt like we needed to switch out everything very, very quickly. In more recent years, our subscriber numbers have increased and it’s really led us to this point where we can’t make the experience great for everyone with our manual programming. We need to scale that. And this is one way that we can do it. We want the experience to be far superior than what [readers] can get from one of our articles found in the wild. Our team has grown so much in the past five years and has added so many different capabilities — personalization being one of them, more editors is another one. [The home screen] is our most valuable platform and the one that we really want to put the most effort into making the best. That has been a shift.

How much of this work is already taking place on the home screen?

Ho: We’ve tried a couple of geo-targeting experiments. We gave readers in California an extended package during [Gov. Gavin Newsom’s] recall election last year. We felt we could add value to one of our largest growing markets, in California. We saw really good results with that. And we wouldn’t have been able to do that with some of the traditional experimentation tools that test headlines. In this case, we are testing a full experience for a certain segment of our readers. Right now, a couple of active [tests] are the “In Case You Missed It” module, right below Opinion. There’s an algorithm that works behind it to showcase some of the breadth of work that we have, as well as amplify some of the strongest pieces. But we’re constantly checking what stories are going there. There are editors behind the pool of stories that are there.

Skog: There is likely a reader for every story we publish and we’re just trying to find those readers. If we know you’re in California, for instance, we can give you more localized content. For the California wildfires, we talked about doing emergency locations to go to that people in New York would not necessarily need to see, but people in California would appreciate. News is magnetic. What we have to work really hard to do is making sure that we surface all of the other amazing journalism that we have throughout the Times for our readers. That’s what makes us special.

Numbers to know

80: Editorial positions that LinkedIn plans to add this year.

9%: Percentage stake that Elon Musk has taken in Twitter.

45: Articles that WNYC removed from its website because they contained plagiarized content.

25%: Percentage share of BBC’s staff that will come from low socioeconomic backgrounds by 2027.

3: Number of podcasts that are departing Patreon to be distributed through Substack.

What we’ve covered

How Refinery29’s Simone Oliver is complementing content with commerce:

• Refinery29 plans to start testing live shoppable video this spring, likely on YouTube at first.
• Oliver joined the Digiday Podcast for a live recording during the Digiday Publishing Summit.

Listen to the latest Digiday Podcast here.

Why TheSkimm is extending its daily newsletter to the weekend:

• TheSkimm has started publishing a new edition of its flagship newsletter every Saturday morning.
• TheSkimm’s weekend edition is created by a team made up of three new hires, a group of writers and other staffers.

Read more about theSkimm’s weekend newsletter here.

Confessions of a former esports journalist who pivoted to marketing:

• The former journalist tired of esports teams taking issue with their reporting.
• One team threatened to sue the former journalist over their reporting.

Read more about the former journalist’s career change here.

Why The New Yorker is using more ‘voice’ in its daily newsletter:

• The newsletter had previously been primarily a collection of links and lacked original material.
• The New Yorker’s daily newsletter has nearly 2 million subscribers and averages more than 1 million daily opens.

Read more about The New Yorker’s daily newsletter here.

What we’re reading

BuzzFeed v. ex-BuzzFeed employees:
BuzzFeed wants the former employees suing the publisher over its IPO to pay half of BuzzFeed’s legal costs, according to Axios.

Spotify v. Parcast Union:
Spotify’s ongoing negotiations with its podcast union is reaching the point where the union’s employees are willing to strike if an agreement is not reached, according to Bloomberg.

Facebook v. misinformation:
Since last October, a bug with Facebook’s content ranking system led to the platform not properly suppressing but instead amplifying posts containing misinformation, according to The Verge.

BBC v. female employees of color:
More than a dozen female employees of color have quit the BBC because they say the British broadcaster continues to favor “white, middle-class and privately educated staff,” according to Variety.