Mark your calendar for Mediaweek, October 29-30 in New York City. We’ll unpack the biggest shifts shaping the future of media—from tv to retail media to tech—and how marketers can prep to stay ahead. Register with early-bird rates before sale ends!
With Google Chrome slated to deprecate cookies by the end of this year, marketers are looking for new signals to find and measure potential audiences. Transaction data—records of what we purchase—is emerging as an alluring alternative.
The promise of transaction data has fueled the retail commerce media industry, which is projected to grow by 28.5% to $59.6 billion in U.S. ad spend this year, per Insider Intelligence.
Retail media firms, notably Amazon and Walmart, can use their customer databases to target ads on their websites and across the internet, and then link those ads with purchases, without relying solely on cookies for closed-loop attribution.
Some think transaction data might become a necessary currency for all media companies, and not just retailers. Only walled gardens with successful commerce businesses can succeed in a post-cookie world because, without cookies, the likes of Meta and TikTok won’t be able to track whether ads on their platforms lead to purchases on brands’ websites.
“If you don’t have a payments business, you can’t build an ads business,” said James Borow, co-founder of ecommerce marketing tech company Market AI.
Others note that transaction data can never fully replace cookies, because these signals are not identifiers. Retail media firms need an email address or cookie in order to know that the person who saw the soap ad bought the soap. Still, in a world where many signals go dark for advertisers, transaction data may emerge as a more critical resource.
The social commerce thesis
Without cookies, even walled gardens will lose their edge for marketers, which need to be able to show that their ads drove conversions, Borow said. When Apple deprecated mobile identifiers on iOS in 2021, Meta struggled to correctly report conversions, causing some marketers to move spend away from the platform. These problems could be exacerbated when third-party cookies are no longer available on the open web.
“You need to have the transaction data to optimize your auction,” Borow said. “Without knowing the types of things that [customers] purchased, it’s very hard to make sure you’re showing the right ad to the right person.”
Since 2021, Meta has begun to regain advertiser trust and combat its signal loss problem via two main tactics.
The first, its conversions API (application-programming interface), helps Meta better target and track ads by having marketers upload their own customer data sets to match with the platform’s first-party signals. But this tactic contains friction and is not as easy as Meta owning transaction data itself, Borow said.
Second, Meta has created artificial-intelligence-powered buying tool Advantage+. These tools have been called out for their lack of transparency, and if buyers begin to lose trust in algorithmic targeting, transaction data could become more necessary, said Shiv Gupta, founder of ad-tech education outfit U of Digital.
“Any ad company that can show ads on its domains and also drive conversions within its domains will be in great shape,” Gupta said. “For this to be a real advantage, a company will need to have significant owned-and-operated inventory and significant on-site transactions.”
Limitations of transaction data
Still, most advertisers won’t abandon set-it-and-forget-it AI buying tools or the behemoth social platforms that invented them. Even if these platforms lose some of their attribution abilities, they are still rich in the ultimate advertising resource: eyeballs.
“What walled gardens have, that retailers and commerce media players don’t, is consumer attention,” said Michelle Dooley, founder at Catalyst Media Consulting, which specializes in retail media and ad tech. “We go to a retail site when we want or need to buy something, but we spend hours on social platforms for entertainment.”
With transactions, you might reach too narrow of an audience.
Nikhil Lai, senior analyst for performance marketing, Forrester
Direct-response advertisers rely on transaction data for attribution. But linking ads to purchase is less important for an insurer or auto advertiser, said Nikhil Lai, senior analyst for performance marketing at Forrester. Even for lower-funnel marketers, most sales still happen in-person, where tracking is more elusive and transaction data is only so useful for targeting.
“By having transaction data in a layer in every buy, you no longer reach people who are light buyers and medium buyers,“ Lai said. “The key performance indicator is new-to-brand customers… [with transactions,] you might reach too narrow of an audience.”
There are already existing use cases for transaction data that might become more in vogue once third-party cookies become obsolete—for instance, using transaction signals as a layer between publisher and advertiser to narrow an ad’s audience to people who have spent more than $35.
Still, using transaction data like this means an additional fee on top of the media CPM (cost per thousand impressions), and targeting that is more contextual and not precise data matching.
“The value there is finding proxies,” Lai said.
