WTF are data collaborations?

This article is a WTF explainer, in which we break down media and marketing’s most confusing terms. More from the series →

There is no shortage of cookieless solutions being touted throughout the digital advertising industry – though the jury is still out as to how good of a replacement they’ll be for third-party cookies. But a relatively new option popping up more in conversations is data collaborations.

During a closed-door town hall session at the Digiday Publishing Summit in March, a publisher exec, granted anonymity under Chatham House rules, spoke about their growing interest in participating in data collaborations with brands. Data collaborations, often abbreviated to data collabs, put the publisher in a more authoritative position, they said, as well as enabled them to scale up the portion of valuable audiences that advertisers could reach. 

“[Advertisers] want to combine data assets to have a better together story, which allows them to think about ways of taking [the brand’s] first-party data and activating it on an unaddressable or on unfindable users that have historically been left behind,” said the publisher during the town hall. 

And largely, the buy-side wants these data collaborations, too, recognizing that they won’t have to part ways with their client’s first-party data once they’re unable to use third-party cookies to target it. 

But what exactly are data collaborations and why are they just now emerging as a viable targeting option in the digital advertising market?

WTF are data collaborations? 

Data collaborations are deals struck between a publisher, brand and (in some instances) retail media networks that are aimed at using first-party audience or customer data from each party to find new prospective consumers that the brand hasn’t otherwise been able to reach. 

“Both sides have valuable first-party data to offer, especially publishers that have login information, and we know that they can target specifically based off of that first-party data,” said Aaron Ledwith, vp of addressable media partnerships at Dentsu, who works with the company’s identity solution platform Merkury. “It allows for brands to see their direct scale on a publisher … and it allows for publishers to sell their insights to brands.” 

OK, but what does it actually look like to operate a data collaboration? 

Ledwith said the typical use case for a direct data collaboration is that a brand will bring its first-party data and Merkury will overlay some of its own custom data. Then, a forecast is created based on the overlaps and projected reach that the combined data sets have for a publisher. After the campaign runs, a report is compiled to determine how the delivered results from the campaign compared to the original forecast. 

That report can be referenced, among other uses, in further planning with the publisher, Ledwith said.

Dotdash Meredith, as an example, approached data collaborations as an extension of the company’s first-party contextual targeting solution D/Cipher. According to Jon Roberts, the publisher’s chief innovation officer, brands’ consumer data has D/Cipher signals layered on top of to reveal insights that both parties wouldn’t have been able to independently determine otherwise.

“Clients know their audience better than we can … We know our audience better than the brands will ever know them. By combining it, we’ve proven we can get unique insights that neither us nor the brands could get to those people on our own. And when we can prove which content will predict this person will become your customer, we can run campaigns that unlock a ton of value. And there is no data leakage because the value is locked in our content,” said Roberts.

Where do data collaborations happen? 

Right now, direct data collaborations between publishers and brands often happen in clean rooms, but they do not exclusively have to occur in those spaces. It depends on the appetite for data privacy and security that publishers and brands desire (or are being held accountable for by regulation).

Andresen said that most of the questions that Gannett gets about testing data collaborations incorporate clean room tech.

Dotdash Meredith’s approach to data collaborations don’t take place within a clean room, however. Roberts said that his team is able to import a brand’s data and match it with its own first-party data on its own tech stack.

Are ad dollars being allocated to data collaborations yet? 

Roberts said that there have been several data collaborations executed with Dotdash Meredith’s clients as his team was building out this capability. But other publishers and buyers said deals aren’t actually closing just yet. 

“A lot of the conversations that we’ve had to date are exploratory of what could we do together and what would it look like if we did this or that and planning to do those approaches together?” said Kelly Andresen, president of national sales at Gannett.

Without disclosing the client’s name, Gannett’s vp of ad innovations, Jeff Burkett, said that one ad agency requested they use a clean room for a data collaboration in 2023 and after significant back and forth between the client, agency, clean room and his team, the deal never materialized.

“In the end, the agency had no idea what to do … and it basically just ended in confusion. Now the conversation is dead,” said Burkett. “It felt like the buy-side didn’t know what they really wanted to do beyond saying we wanted to use it.”

Seth Hargrave, CEO of ad agency MediaTwo said that his company has not started transacting with any data collabs, but it is something that he’s monitoring closely. “The ultimate result could be that we will see these data collabs emerge essentially as almost a standalone publishing entity, because that data is going to be so valuable to buyers,” he said.

Clean rooms have been around for several years. Why are data collaborations just being talked about now? 

In part, it took time to build out the technological capabilities to do these data collaborations safely and securely.

“It’s been an evolution. We’re now in that second stage of practice, not just theory,” said Alex Reeder, vp of audience and data solutions for Havas Media Network.

Dotdash Meredith’s data collaboration offering took upwards of a year to build out, said Roberts, given the complexities of onboarding data and matching it with the company’s D/Cipher segments. But now, he said there are enough case studies of completed tests with brands that the process is repeatable for other, interested brands.

On the other hand, interest in data collabs has increased as cookie deprecation on Chrome progresses. 

“Clean rooms are not new, but the interest in them and the exploration around them has definitely picked up,” said Andresen. “I would attribute this directly to being aligned with the deprecation of the third-party cookie. There’s nothing pressing that advertisers have to do right now to test out [data collaborations], so I would expect … that we would continue to see that pace of exploration [increase at the rate of cookie deprecation].” 

Like any cookiepocalypse preparedness plan, most agree that it’s better to know which options work for media plans now than wait and test when there isn’t a third-party cookie back-up to cushion missteps.

“If cookie deprecation does go forward, those that have already started testing and utilizing these types of solutions will be better off because they already have a way to scale and activate against their intended audience. And they already have an understanding of the publishers who can help them do that,” said Ledwith.

“We see data collaboration as one of the tools to help solve signal loss. Cookies might be going away, but publishers are collecting much more authenticated and consented data. So brands are figuring out maybe it’s not a matter of match rates. Maybe it’s going back to reach rates,” Reeder said.

Meanwhile, Roberts argued that data collaborations should be happening regardless of third-party cookies being available to advertisers. “We’re not waiting for cookie deprecation because [data collaborations] already solve 50% of the internet that was previously untargeted. If we can get [our advertisers] comfortable with that … then that gives everyone a really stable basis because this way of targeting does not change when cookies deprecate,” he said.

Do data collaborations only work with authenticated audience data? 

The industry seems to be split on this one. While buyers have more confidence in who a prospective customer is, publishers said this can limit the benefits of a data collaboration.

Reeder said he currently has “more confidence in the authenticated, first-party consented collected experience,” versus unknown audience data. Granted, there will likely only ever be a fraction of a publisher’s audience that falls into the authenticated camp, but Reeder said even incremental data added to a brand’s customer data set is helpful.

Dotdash Meredith’s Lindsay Van Kirk, svp and gm of D/Cipher, said that her team doesn’t limit data collaborations to only authenticated audiences because, “the value is not just in the collaboration, we’re actually providing them the opportunity to extend the applicability of [their data] … Don’t limit yourself by only matching to an identifier,” which is always going to be limited in scale.

“They come for the data, but they stay for the insights. And that helps us build a much more strategic, deeper partnership than just that tactical data transaction,” said Roberts.

More in Media

Inside The New York Times’ plans to correlate attention levels to other metrics

There’s a lot of buzz around attention advertising right now, but The New York Times is trying to stay grounded even as it develops its own plans.

Why publishers are preparing to federate their sites

The Verge and 404 Media are exploring the fediverse as a way to take more control over their referral traffic and onsite audience engagement.

Why publishers fear traffic, ad declines from Google’s AI-generated search results

Some publishers and partners hope for more transparency from Google and other AI companies related to AI-generated search.