Cambridge Analytica's effectiveness called into question despite alleged Facebook data harvesting
Cambridge Analytica’s use of “psychographic” voter information was neither novel nor measurably effective, experts in the field tell NBC News.
In this photo illustration the logo of the strategic communication company "Cambridge Analytica" is seen on the screen of an iPhone in front of a computer screen showing Facebook logos on March 20, 2018.Chesnot / Getty Images
Breaking News Emails
Get breaking news alerts and special reports. The news and stories that matter, delivered weekday mornings.
WASHINGTON — Cambridge Analytica’s use of “psychographic” voter information to target people based on opinions, attitudes and personality traits was neither novel nor measurably effective, experts in the field tell NBC News.
But it’s not clear that the $15.4 million Trump and other U.S. political entities dished out to Cambridge Analytica in the 2016 election cycle was money well spent.
“What they’re doing is bullshit, basically,” said Antonio García Martínez, a former Facebook employee who helped build the social network’s ad-targeting operation. “This has been tried before. There’s no reason to think it’s particularly powerful.”
In theory, Cambridge Analytica could predict and influence voters' behaviors by delving into personal details shared by Facebook users through an "OCEAN" personality survey, which measures openness, conscientiousness, extraversion, agreeableness and neuroticism. The idea behind "psychographic" targeting is to use those personal details to create advertising that is more likely to influence voters to choose a certain candidate or to come out to vote.
Martínez said that the targeting by Cambridge Analytica might have been considered good by political advertising standards.
“The reality was that political ad targeting in the past wasn’t very good," Martínez said. “I think what’s happened is they just caught up."
In a presidential election decided by roughly 80,000 votes in three states, many factors could be considered potentially determinative. But Republican and Democratic operatives familiar with data science and the type of information reaped by Cambridge Analytica say there's scant evidence that the firm was effective — or that it provided information more valuable than the basic voter-file data the Trump campaign could have gotten elsewhere.
Let our news meet your inbox. The news and stories that matters, delivered weekday mornings.
That is, they don't believe the election turned on Cambridge Analytica's involvement.
"Any data is only useful in so far as it can help me build a more predictive model," said a leading GOP data scientist who spoke on the condition of anonymity because of the sensitivity of the topic. "Whether or not someone likes Taylor Swift or did a keg stand in college is far less useful to me than whether or not they voted in the Republican or Democratic primary in previous years. And until someone proves to me that OCEAN score targeting truly can be used to behaviorally persuade a voter, I’ll remain skeptical of the value."
Eric Wilson, a GOP digital political operative, said that even if the personal details gleaned from Facebook were more potent than other data, a campaign would still have to figure out how to tailor ads to voters.
"The psychographic data is only as good as the creative you’re able to generate," he said.
And a Democratic operative who works in the digital realm said claims of Cambridge Analytica's effectiveness in 2016 are very much overstated.
But Christopher Wylie, who helped found the company and served as its director of research until 2014, argues that its work is subtle enough that it might not be noticed.
"A lot of what information operations is, is about warping people's perceptions," Wylie told NBC News. "This is a company that was birthed out of a military contractor. The premise is not necessarily to do standard political messaging where people are aware that I am trying to be convinced of something. It is to change what people think and perceive about what is real."
Trump's campaign spent $5.9 million on Cambridge Analytica services; Ted Cruz's presidential campaign shelled out $5.8 million to the company; and Make America Number 1, a superPAC funded by billionaire Robert Mercer, who backed Cruz in the GOP primary before switching to Trump in the general election, paid the firm about $1.5 million.
Wylie said that Steve Bannon, a then-Mercer ally who later became the chief executive of Trump's campaign, was particularly interested in what the company was offering.
"Steve wanted weapons for his culture war," Wylie told ITN Channel 4. "We offered him a way to accomplish what he wanted to do, which was change the culture of America."
Cambridge Analytica has rejected Wylie's assertions, and a spokesman for the firm said he is engaged in "a malicious attempt to hurt the company."
The Trump campaign and Cambridge Analytica actually had an underwhelming digital operation, said Jessica Baldwin-Philippi, an assistant professor at Fordham University who has studied the digital efforts of political campaigns.
"The idea that they had a robust data campaign is just not true," she said. "They were doing very best practice stuff on the Facebook side, but they weren't doing this universally. It wasn't deep or robust."
She agreed with the political operatives that there is little evidence to support claims that Cambridge Analytica helped swing the election.
"Time and again, studies have shown us that the most persuasive targeting metrics are not crazy-specific microtargeting data but public-record voting history data and geography," she said.
CORRECTION (March 21, 2018, 12:25 a.m. ET): A previous version of this article misspelled the last name of an assistant professor at Fordham University. She is Jessica Baldwin-Philippi, not Baldwin-Phillipi.