With additional reporting by Alison Weissbrot.
Sizable brands and top-tier publishers can use their clout to pressure media platforms into sharing more measurement data – but what are advertisers actually getting?
Just enough to keep them spending, said Chris Kane, founder of programmatic consultancy Jounce Media.
“Walled gardens will do the minimum acceptable thing so that advertisers don’t break up with them,” Kane said. “The fact remains: No one has the full picture on measurement – and advertisers are unlikely to ever get it.”
What’s on offer?
Google and Facebook are each responding to advertiser demands for more data.
Facebook does data-sharing deals on the DL with large marketers that push for it.
In so-called “clean rooms,” for example, advertisers can compare their first-party data with impression-level Facebook campaign delivery data using laptops that have never touched the internet. Facebook also allows certain large advertisers to create a private instance on its server to run advanced analytics.
Facebook told AdExchanger it’s mulling how to offer more data access at scale, although what form that will take is still unclear.
Google is already taking a more productized approach with Ads Data Hub, its cloud-based system for the pixel-free measurement of YouTube and media bought through DoubleClick and the Google Display Network that’s set to come out of beta this quarter.
The Google strategy is centered on “building infrastructure,” said Dan Taylor, managing director of global display at Google. “We’re not investing in anything bespoke with particular agencies or customers.”
But, according to some insiders with direct knowledge, Google will enable custom measurement setups for select brands. This isn’t something Google publicizes or does regularly.
“If we need to do something important for an account, Google will consider it,” the head of media at a large brand told AdExchanger.
The third elephant in the room, Amazon, isn’t offering anything along these lines, at least for now. Advertisers have no access to Amazon audience segments outside of Amazon’s advertising systems.
Another brick in the wall?
But despite sharing more data and the overtures to openness, a contrarian view is that the garden walls are actually going up rather than coming down.
Unconventional data-sharing programs are a) rather rare – which makes it difficult to get a read on how often they happen, and the platforms aren’t talking – and b) usually about pulling more data into the platform, not the other way around, Kane said.
For example, Facebook is running an alpha test to give advertisers anonymized info about people who have made purchases based on log-level data. The catch? Facebook has to put a pixel on every purchase page. Then it notifies the advertiser what’s being driven by Facebook and what isn’t. (Google offers something similar and more fully baked in the form of an AdWords conversion tracking tag.)
“You basically have to expose your entire P&L to Facebook if you want to be part of this – and all of our competition is spending on Facebook,” said a marketer who was pitched on the solution, but didn’t bite. “When Facebook approaches you with something, it’s usually more advantageous to Facebook than to you.”
Is it worth the effort?
Another problem is, although niche data-sharing arrangements hold the promise of better attribution, they don’t solve for cross-channel measurement – and that’s what advertisers are really looking for. As another marketer put it: “What’s the value of this data if I can’t trend it? It’s just a snapshot in time.”
These deals also require a lot of work and man-hours from multiple stakeholders across the brand and agency, from legal and IT to media planners, data scientists and engineers.
And for some, the juice isn’t worth the squeeze.
“You can bribe the platform to do something like this, but you’re using a lot of agency time and your own team’s time to generate data for a single campaign,” said the head of integrated marketing at a large CPG brand. “The value of it all collapses under its own weight.”
But for advertisers with the appetite, strict terms govern what data they get back from the platforms and in what form. There’s also usually a spend commitment or a minimum number of impressions, which is why only the big guys qualify.
From there, depending on the setup, first-party data is mingled with anonymized campaign or audience data on the platform side within a closed-off environment, whether that be in a clean room or in the cloud.
The data that advertisers get back can be used for attribution, measurement, modeling or audience discovery.
It’s a heavy lift, it can take months to get through legal, the data that comes back is only from one campaign on one platform and the setup is far from scalable. But, then again, scaling a solution like this isn’t something the platforms would even attempt if they weren’t under pressure. These deals are an appeasement, and the advertisers that partake go in with their eyes open.
“We’re only getting measurement for Facebook or for Google in a silo for one period of time for one type of spend – and it’s a pain in the ass to execute,” said one large marketer. “Even so, that makes our internal attribution model smarter. It’s all incrementally helpful.”
Open to openness?
But how helpful can a platform be without giving away the farm and running roughshod over user privacy?
Facebook is pondering how to answer that question.
Although personally identifiable information will never be on the table, tackling cross-platform measurement is a top priority for Facebook in 2018.
“Marketers don’t have a real easy way to compare apples to apples across all the platforms,” acknowledged Carolyn Everson, Facebook’s VP of global marketing solutions, at AdExchanger’s Industry Preview in January.
The scaffolding to provide better measurement is there. Facebook submitted to a Media Rating Council audit last year, as did Google, and both maintain numerous third-party measurement partnerships, including with Nielsen, comScore, Moat, Integral Ad Science and DoubleVerify.
An independent third-party will be the key to cross-platform measurement, Brad Smallwood, Facebook’s VP of marketing science, told AdExchanger.
The mechanisms are there. The tricky part, Everson said, is figuring out how to strike the proper balance between data sharing, consumer expectations and privacy regulations, like GDPR in Europe.
“It’s easy to say, ‘We need more data,’ but you actually have to get underneath this and realize that it’s complicated,” she said. “It’s like peeling an onion and we have to do things in a very privacy-safe way.”
But hope springs eternal
For now, advertisers do what they can to hack together siloed data while the anointed few strike niche data-sharing deals with the platforms that offer them.
It’s “not quite the interconnectedness that we would like” between the platforms, said Oscar Garza, global director of media activation at Essence. “But it’s a step in the right direction.”
There’s value in more attribution data from Facebook and Google, even if it’s not cross-platform, he said.
“We hope for a time in the next year or so for attribution to be applied in more places – and if [the platforms] believe in their inventory and the formats they offer, those will yield positive results for them,” Garza said. “Major players are making real, conscious efforts to build those bridges.”
