‘Psychographic’ wizardry or psychobabble?
Some Republicans heard Cambridge Analytica’s election data pitch and found it to be digital snake oil.
WASHINGTON — When Cambridge Analytica officials met GOP strategist Mike Murphy at the start of the 2016 presidential campaign, their pitch was slick and full of swagger — but after a little probing, Murphy ultimately found it to be full of nonsense.
At the time, a meeting with Murphy was in high demand. He was heading Right to Rise, a PAC that was practically printing money, ultimately raising $118 million in its unsuccessful effort to elect Jeb Bush. Murphy’s team concluded that Cambridge Analytica had nothing to offer other than hype.
Disclosures this week that the firm used massive amounts of data from Facebook to develop profiles of American voters to help Donald Trump’s campaign have generated a public image of Cambridge Analytica as devious masterminds of electoral manipulation. A secretly videotaped interview in which the firm’s CEO lays out how it could help clients blackmail rivals only fueled that perception.
But some GOP campaign professionals, data scholars and former clients offer a very different tale.
“They were telling me what they had to sell was more advanced than anything I had ever seen before,” Murphy said of the now-embattled firm’s promise to use
“psychographic profiling” that would cycle the tastes and interests of millions of voters into algorithms that then target them with tailormade, highly persuasive digital ads. “They were just throwing jargon around,” he added, recalling that the firm also claimed it did topsecret work for the military.
Whether the firm, even with the massive amounts of Facebook data it is accused of improperly mining, was the digital Svengali it marketed itself to be is very much in doubt. For some clients, Cambridge Analytica was more a conduit to endearing themselves to its billionaire financial backer, Robert Mercer, who was also a major contributor to conservative campaigns.
“The thinking was if it gets Mercer to fund your super PAC, then it’s worth it to use this firm,” said Luke Thompson, vice president for politics and advocacy at the Republican analytics firm Applecart. But he said Cambridge Analytica’s claim that it has reinvented political persuasion is based on “ludicrous assumptions and leaps of faith.”
Thompson was helping lead data efforts for the National Republican Senatorial Committee when the firm pitched its products in 2014. Thompson, a former academic who taught at Yale University, found the claims it was making absurd. He said the psychographic mapping that it promised could be done on a large scale struck him as improbable, and of limited value even if it did pull it off.
“It was all based on this pop psychology BS, and even if you could do it, it would add only the most marginal value to a campaign,” he said. “Nobody was asking them the most basic data-sourcing questions.… If you want to know what people think about politics, why would you do these surveys where you ask them about things like antique cars? Why not just ask them about politics?”
Such criticism of the company, which is common among GOP analytics leaders, was muted by its connection to Mercer, a hedge fund billionaire who built his fortune creating a formula of algorithms and equations for investing that only a few math geniuses understand.
“I don’t understand how a man as smart as him put all this money into this black hole of nonsense,” Thompson said. The firm’s profile also got a boost from its connection to former Trump campaign chief and senior advisor Stephen K. Bannon, who had been a vice president there.
Cambridge Analytica did not respond to questions about Mercer’s involvement or the firm’s track record.
But company leaders have frequently touted successes, claiming its proprietary techniques helped propel various prominent politicians to victory. They point, for example, to the triumph of Trump and early primary wins by Republican Ted Cruz, the Texas senator.
The Cruz campaign initially embraced the firm’s narrative that it was using cutting-edge technology to engage and persuade voters in ways that were transforming politics, but those who worked for Cruz later cast doubt on Cambridge Analytica’s effectiveness.
The firm’s salesmanship persuaded Todd Wilcox to sign up for its psychographic profiling when the conservative super PAC he was running, Restore American Leadership, was aiming to elect business leaders and veterans to office. Wilcox was unimpressed.
“They said there was this whole predictive analysis they could do, but we didn’t see any evidence it worked,” he said. Wilcox also recalled Cambridge Analytica touting how it could engage and persuade voters by using certain words and colors in campaign graphics. “We stopped using them,” he said. “We saw no evidence that different color shades and verbiage motivated people in a different way.”
The firm earlier disputed in a statement to the Wall Street Journal the way Wilcox describes the relationship, saying it did only digital marketing and web development work for his group, not deep analytics.
A conservative organization called the Market Research Foundation was similarly uninspired when it partnered with SCL Group, the company that would become Cambridge Analytica, in a voter persuasion effort leading up to the 2013 Virginia governor’s race. “We would not recommend working with them on any project,” the foundation wrote in a memo that concluded the firm’s psychographic targeting effort was a failure.
When Cambridge Analytica Chief Executive Alexander Nix later published a 65page white paper touting the success of psychographic profiling in Virginia, the foundation felt compelled to respond with its own memo warning other Republicans that Nix’s pronouncements did not track with what actually happened. “MRF was astonished by several of the claims made,” the foundation’s memo said.
Those who have worked with or studied the firm are amused to see Cambridge Analytica now cast as an evil genius with the power to manipulate an election. While the transgressions the firm is accused of carrying out are serious, critics who understand its work best say the firm just wasn’t equipped to do that much damage.
Eitan Hersh, a Tufts University professor and author of “Hacking the Electorate,” suggested that all the attention Cambridge Analytica is getting in the controversy over its alleged misuse of Facebook data is misplaced. Like many others interviewed, he said the firm sells snake oil. “To me, the story is 99% about Facebook and 1% about Cambridge Analytica,” he wrote in a direct message over Twitter.
Yet as lawmakers dig into how the Facebook data were used and a special counsel investigates whether there was coordination between the Trump campaign and Russian operatives to influence the 2016 election, it’s difficult to look away from Cambridge Analytica.
On Tuesday, its board suspended Nix over, among other things, his comments to a British reporter posing as a potential client from Sri Lanka. Nix suggested sending “girls around” to the home of a prospective client’s rival, ostensibly for a blackmail scheme.
There was also undercover video of Nix claiming Cambridge Analytica played an outsize role in Trump’s victory. “We did all the research, all the data, all the analytics, all the targeting,” Nix said in the video captured by Britain’s Channel 4. “We ran all the digital campaign, the television campaign, and our data informed all the strategy.” Many in the Trump campaign disputed that.
As he tries to make sense of it all, David Karpf, a George Washington University professor who specializes in digital advocacy, says Cambridge Analytica is looking less like a mischievous genius than a mischievous amateur. He compares its alleged Facebook intrusion to burglars who set out to rob a vault full of diamonds and end up hauling home a bag full of worthless cubic zirconia.
“But even if the robbers only got junk out of it, it is still a crime,” he said. “And it is still concerning that the vault was so easy to rob.”