Far from harming the business outcomes of advertisers’ digital marketing, the end of third-party cookies and current audience targeting practices may actually make things better, according to Dr Augustine Fou.
The practice of collecting huge volumes of data via cookies – hundreds, sometimes thousands, of trackers may be loaded into a single webpage by adtech companies – has been designed to create audiences and improve ad targeting.
Writing for WARC, Fou, a cybersecurity and ad fraud investigator, notes academic studies that reveal just how poor ad tech data quality can be. If the accuracy of targeting gender is 42%, as one study found, that’s worse than using no targeting at all.
“If even getting the gender correct is difficult, how accurate would other, more complex audience segments be when created from website visitation data collected by privacy-invasive trackers?” he wonders.
His conclusion is that the lookalike audiences that advertisers are buying based on inferences made about people visiting various sites are made-up ones. “On average those audiences are about 35% better than no targeting at all – the natural audience.”
In fact, the accuracy of audience targeting varies widely, from 7% to 77%; as well as initial data flaws and incorrect inferences, there’s the often unconsidered matter of bots which “pretend to be any desirable audience segment in order to earn more money for their creators”.
Thus bots are programmed to visit medical journal sites to make themselves appear to be doctors and in this way they earn higher CPMs from pharmaceutical company ads seeking to target the audience segment of doctors.
“The same bots can also be programmed to visit pages with backpacks and uniforms on Amazon, to pretend to be back-to-school ‘intenders’, a highly desirable audience segment for many advertisers,” Fou adds.
And if that’s not enough food for thought, “marketers should also carefully consider whether they are actually getting anything for the extra costs they paid for audience targeting,” he advises. “By most accounts, behavioral targeting is not worth the extra expense.”
He cites figures from the same academic study to show that advertisers “paid 2.5X more costs and got at best 30% incremental outcomes compared to ‘spray and pray’” – or, effectively, a negative ROI.
Fou’s conclusion is that “Advertisers will achieve more business outcomes from digital marketing by focusing on incrementality, slashing audience targeting costs, cutting the adtech expenses that don’t yield more business outcomes, and moving money away from behavioral targeting, microtargeting, and hyper targeting, and back to the basics of marketing”.
For more details, read Dr Augustine Fou’s article in full: Audience targeting: Why losing cookies may not matter.
Sourced from WARC
