After years of hand-wringing over the interminably slow death of third-party cookies, the advertising industry has started to have a change of mind, if not a change of heart on the matter. An acknowledgment that cookies have rarely provided much quality in terms of actionable insights on driving consumer behavior has steadily taken hold.
First-party data has always commanded respect for its direct access to valuable consumer information. From media consumption habits to product interests, it was commonplace to view it as something brands hoarded unfairly. Cookies, by contrast, are (still) widely available across the ecosystem. But that easy scalability also came at a price; cookies simply weren’t that effective.
“Cookies weren’t really designed to help us understand consumer behavior across multiple devices in order to understand consumers’ potential needs and wants,” writes Dominic Charles, managing director, Audience Intelligence and Marketing Science at media agency Wavemaker.
As Charles notes, cookies were initially used to allow websites to recognize returning users so that they weren’t forced into the frustrating act of entering their login information every time they clicked in to visit.
“As a result, data collected was incredibly fragmented and the solutions to solve this problem pretty much doubled down on collecting even more individual-level data without consent,” Charles notes.
The consensus is that first-party data creates a high quality seed audience based on deterministic data — information a user shares with an entity through services such as a subscription or loyalty program — rather than cookies’ probabilistic data, which is essentially a guess about future actions based on an aggregate of past purchases and website visits.
Even with traditional cookie-based audiences, lookalike models are still used to expand the reach beyond the seed. This takes any inaccuracies from the probabilistic cookie audience and exponentially multiplies them.
Deterministic first-party data, on the other hand, can create a seed audience worth modeling. The advancements in machine learning and methodology also allow for smarter lookalike models that are continually learning from fresh first-party data, compared to cookies which quickly become stale and inaccurate based on the nature of web browsers.
As brands and media companies anticipate a decline in search traffic due to artificial intelligence-generated answers replacing the endless “blue links” in response to queries, the quality of data depends on the existing connections brands have with their current workflows that strengthen their insights into consumers, said by Federico Salvitti, VP of Growth US at end-to-end advertising management platform MINT. He points out that AI enables real-time adjustments and steady, incremental improvements.
“AI has revolutionized advertisers’ access to actionable data by enabling closed-loop workflows that allow for iterative improvements and real-time adjustments, enhancing marketers’ ability to learn and convert,” Salvitti said.
A significant challenge in the industry has been the timely measurement of campaign performance – crucial for taking action quickly enough to positively influence the campaign.
“The biggest challenge in advertising is that KPI results arrive too late, which confines insights to only be used at a campaign strategy level,” said Salvitti. “KPI analysis should happen daily, and your ability to make changes to your trajectory should be as simple as clicking a button that says ‘apply changes.’ Effective data utilization requires seamless workflows that integrate media mix modeling, activation, reconciliations, and payments, allowing for instantaneous adjustments.”
This vision of the future plays nicely into AI’s strengths - taking a constantly flowing, complex set of performance metrics, looking for patterns, and making continuous optimizations for optimizations. Yet this opportunity is completely reliant on the availability of high-quality, consented data and persistent identifiers.
“This ultimately creates more opportunities for ethical advertising,” said Natasha Tibbetts, VP Media at ad agency Swell Media. “Only then are the massive amounts of data from compliant sources able to lend targeting to more refined and precise audience segments, visibility into audience and inventory performance for agile optimizations, and attribution modeling for conversions and sales accuracy across multiple touch-points.”
.png)
