Podcast: The Very Serious, Not At All Overblown Facebook Scandal

Welcome to AdExchanger Talks, a podcast focused on data-driven marketing. Subscribe here.

True or false: While Facebook was looking the other way, Cambridge Analytica exploited easy loopholes in its data policies to steal user data and swing a US presidential election.

In this week's podcast, Gartner Research VP Martin Kihn picks apart the assumptions underlying the latest Facebook backlash: namely, that the data scraped by Cambridge Analytica had a material impact on the election outcome.

Whatever Cambridge Analytica's level of data sophistication, and it seems the firm is pretty smart, Kihn says, "Even if you have the perfect insight into somebody … how does that then translate into designing the perfect creative and getting them to change their vote? It's a leap too far."

Kihn analyzed the available research on political advertising effectiveness and concludes: "You have to spend a lot, you have to do it right by the election, and you can maybe see a bump, but the evidence that you can change somebody's vote is very slim."

The reason: "externalities." In the language of mix modeling, externalities are simply external (non-paid media) factors that influence a purchase. In politics they are legion, even more so in the 2016 cycle. Candidate Trump's earned media impressions blew the roof off any other. Put another way, externalities dwarfed paid impressions.

"He just was on all the time," says Kihn. "The impact of paid advertising on this campaign … would be absolutely at the margins. To overreact to Cambridge Analytica is ridiculous."

Also in this episode: Customer data platforms, Marty's "Taylor Swift" theory of branding, subliminal advertising.

1 Comment

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>