AI is helping marketers master the open web — even as it rewrites the rules
Illustration by Robyn Phelps / Getty / Shutterstock / The Current
AI is increasingly helping marketers extract more value from their first-party data and navigate a fragmented media landscape, even as it threatens the very ecosystem it draws from.
As AI chatbots begin answering more user queries without directing people to publisher or brand sites, advertisers could risk losing access to key sources of first-party data due to declining web traffic.
In response, some marketers have been rethinking how to better leverage the first-party data they already own — using AI to do everything from building high-propensity lookalike audiences to optimizing bids in real time and predicting which media channels are likely to drive the most efficient outcomes.
“AI has significantly transformed the way our agency approaches audience targeting and media buying,” says Amy Hilton, senior programmatic trader at Doe-Anderson in the U.S.
From Germany to Australia to the U.S., marketers are leaning into AI-powered audience modeling to help them work towards sharpening their targeting and reducing acquisition costs. They’re also using AI to navigate the maze of media channels, stitching together omnichannel buys across connected TV (CTV), audio and display.
First-party data meets AI
Sky Deutschland used its first-party data to build a high-intent seed audience — such as users who had added a subscription package to their cart — and activated it across CTV campaigns using AI-powered buying in The Trade Desk’s platform.
Working with agency EssenceMediacom Germany, Sky created lookalike audiences that outperformed traditional demographic and interest-based ones. The result: seven times higher audience relevance resulting in an 84% lower cost per action.
“The use of first-party data within our CTV campaigns has proven to be valuable. We reach a high-quality target group and can still achieve a high reach, thanks to predictive models,” says Katharina Weiß, senior manager of programmatic and data, digital media channels team at Sky Deutschland.
They’re not alone. Cashrewards, an Australian cash-back platform, used its first-party data to build a high-value seed audience and layered in Zenith’s Imagine Consumer Panel data to create lookalike segments. These were activated across broadcast video on demand (BVOD), audio and display.
The strategy paid off: The campaign reached seven times more target consumers than its original goal and slashed cost per acquisition by 73%.
One AI to rule them all
While some marketers are using AI to refine who they reach, others are using it to rethink how they reach them.
Doe-Anderson, a U.S. agency, adopted an omnichannel strategy for a leading destination brand aiming to boost online ticket sales.
The agency consolidated its media mix across CTV, audio and display — retargeting households already exposed to one channel by introducing them to another. This cross-channel targeting saw 30% of the audience engaging with two or more channels, with the campaign driving a 36% decrease in cost per acquisition.
“When we combine [The Trade Desk’s] Koa with first-party data, such as purchase history or website behavior, we consistently see strong performance, often matching or even exceeding our retargeting audiences but with greater scale,” Hilton says, referring to The Trade Desk’s AI engine.
What’s ahead for AI in media buying
Former WPP CEO Sir Martin Sorrell caused consternation when he recently predicted that AI would “completely gut what goes on in media planning and buying,” adding that “it’s all going to be machine-driven.”
But programmatic traders actually running campaigns tell a different story — one where humans still lead, aided by sophisticated AI tools.
“While AI tools currently support optimization around bids, audience selection and pacing, I believe their role will soon expand to become more integral to precampaign strategy development,” Hilton says.
Hilton envisions future capabilities such as data-driven budget allocation across channels like CTV, display and audio; AI-suggested targeting strategies and frequency caps aligned with campaign objectives; and predictive performance modeling for various campaign parameters — all “before a dollar is spent,” she adds.
“These advancements will enable traders and strategists to make smarter, faster decisions, reduce guesswork in the planning phase and create media plans that are both more efficient and more scalable.”
The Current is owned and operated by The Trade Desk Inc. This information is provided solely for background and is not a representation or guarantee of any future performance.