How to manage the tradeoff between personalized advertising and privacy is a familiar debate.
Whether personalized advertising – or advertising, full stop – works is a less common direction. But in a debate organized by Facebook (surprise!) between IAB CEO Randall Rothenberg and “Subprime Attention Crisis” author Tim Hwang, Hwang challenged the widely accepted notion that data-driven advertising works.
“We may be in an industry-wide situation where correlation is confused with causation,” Hwang said, a point-of-view influenced by researcher Nico Neumann. “For all this machine learning and AI-driven targeting we’re often finding consumers who would have bought these products anyway.”
So, are marketers just lighting hundreds of billions of dollars on fire?
Maybe. Hwang compared digital advertising to the housing crisis of 2007, when leading financial experts turned out to be utterly wrong. “That marketers ‘feel’ it’s working raises red flags to me,” he said.
The counterpoint from IAB CEO Randy Rothenberg was unequivocal: “Economists are overall in agreement that advertising works,” he said.
While Rothenberg was strongly in favor of personalized advertising, Hwang questioned whether personalized ads were worth it.
The “grand bargain” is that if you give up your data, ads will be relevant for you. Although people often use personalized ads and relevant ads interchangeably, Hwang noted a subtle difference. These systems generate personalized ads, he acknowledged. “The question is if they generate relevant ads.”
While Hwang said contextual advertising could avoid massive consumer data collection, Rothenberg warned that the move would be disastrous for small businesses, while consolidating spend with the larger players – like debate host Facebook.
“When you reduce the ability for publishers to deliver relevant ads to target groups based on behaviors, you start rewarding the largest, scaled players,” Rothenberg said.
But the ease of data collection and big data analytics have led companies to amass huge quantities of user data, most of which they don’t need. While consumers are most afraid of their data being used for identity theft or being compromised with malware, “they are worried that advertising helps create and foment those vulnerabilities,” Rothenberg said.
Besides the data concerns, people may not have a choice when it comes to trading their data and attention for content and services.
Rothenberg also outlined the privacy paradox, where some people say that they value their privacy but choose free services.
But Hwang said consumers “feel that this bargain isn’t working.”
“That a lot of data is being given up, and there is a not a lot of return being given back for the consumer,” he added.
When a big chunk of consumer disappears next year along with the end of the third-party cookie, some of these questions about the value of personalized advertising will be answered.
Academics couldn’t ask for a better natural experiment to prove not only whether advertising works.
“It’s an experiment that frankly an academic is not going to be able to do,” Hwang said. “It is the big moment.”
