What We Learned From a Former Cambridge Analytica Exec
Former Cambridge Analytica director Brittany Kaiser talks Facebook, data mining, and using social media profiling for political persuasion.
By Rob Marvin
Brittany Kaiser recently appeared before the UK parliament to give evidence on her former employer, data firm Cambridge Analytica. A day later, she sat in a hotel conference room in Manhattan speaking to a roomful of reporters about data privacy.
The irony is not lost on her. Kaiser worked for Cambridge Analytica and SCL Group (Cambridge’s parent company) in various roles from December 2014 to March 2018. In that time, she served as Special Advisor, Director of Program Development, and Business Development Director. But she is now speaking out alongside whistleblower and former Director of Research Christopher Wylie.
Kaiser is deeply embedded in the blockchain space, serving as a co-founder of the Digital Asset Trade Association and Executive Advisor of IOVO, the new decentralized data ownership platform that this week’s press conference was ostensibly about. The media, however, was there to talk to Kaiser about Cambridge Analytica.
Since leaving the firm, she has also started the #OwnYourData campaign to lobby Facebook and other social networks to change their data policies. Kaiser knows that going from mining people’s data to helping them protect it may seem hypocritical, but her professional 180 is due to the fact that she understands exactly how the industry works.
“To be honest, for many years I never questioned it. That’s the way the political system works; it’s the way advertising works; it’s the way every industry that exists on digital communications works,” said Kaiser. “The data crisis around the world is letting us raise ethical questions about how we start changing that.”
After the press conference, I sat down with Kaiser to discuss how Cambridge Analytica worked on political campaigns in the US and around the world, how the firm created those viral quizzes that scraped your data, and what happened inside Cambridge Analytica after the scandal turned the company upside down. Here’s what we learned.
Kaiser said her career in data analytics began 11 years ago, when she worked on the Obama campaign’s new media team. “The use of data for political purposes wasn’t invented by Cambridge Analytica. It started when I was on the Obama campaign in 2007–2008,” she said. “We invented social media strategy. Since then it’s been a journey,” she said.
After that, she started doing consulting on social media strategy for charities, NGOs, and political campaigns, including Amnesty International. She kept running into the SCL Group consulting on the same things, and her official parliamentary testimony states that she met former Cambridge Analytica CEO Alexander Nix while working with Democrats Abroad in London in 2013. She joined SCL Group in late 2014.
How the Firm Devised Those Viral Quizzes
Kaiser said the creative, psychology, and data science teams would work together to design the social surveys and personality quizzes that served as the foundation for much of the improperly obtained data used for targeting. The questionnaires were designed to figure out which words, concepts, or images would resonate with people.
“In order to create predictive algorithms, you need to have a training set. So, that training set is created through our quantitative surveys. Those surveys need to include either basic market research questions or basic political polling questions, which might be added to get your opinion on a brand or an issue or a candidate,” said Kaiser.
It wasn’t just Facebook, either. Cambridge Analytica’s digital team built a product stack targeting more than 30 “unique inventory sources,” she said, meaning social apps, search engines, and the web.
“We also specialize in psychographic questions. Delving into different aspects of your personality. Questions like, do you get along well with children? Do you believe in the importance of art? Do you see yourself as a leader in society? Do you give back to your community? Those help to understand how an individual sees the world. That completely frames the messaging that will then be used.”
Kaiser told Parliament that while she never worked on the research side directly, she was aware of a wide range of surveys created in this manner. Two specific ones she recalled were a “Sex Compass” quiz to find out your personal preferences, and a “Music Personality” quiz.
Kaiser said she didn’t recall any other specific surveys, but she gave general details about how these kinds of quizzes were phrased.
“I used to always phrase examples in my sales meetings about what those quizzes were. At the time, there were all these viral quizzes, like ‘What Disney Princess Are You?’ Everybody seemed to be taking that, or stuff like ‘What City Should You Be From?’ or ‘What Country Should You Live In?’
“Those are the types of questionnaires that were masked in order to undertake data collection by companies all over the world,” said Kaiser. “Cambridge Analytica didn’t do any of the ones I just mentioned, but in order for people to understand what I was talking about, I would make reference to the ones that were most popular at the time on Facebook. You could collect individual data sets that were obtained from users signing in with Facebook, and those were more important than the actual answers to the questions.”
Beyond Social Media
In her various business development and sales roles, Kaiser spent most of her time meeting and pitching clients for the firm to work with. She talked about data analytics work Cambridge Analytica did across a host of different industries.
“For a long time I was concentrated on building out the commercial division of our company. We would be meeting with consumer packaged goods companies, automotive, retail, fashion, etc. These were the types of companies that are used to collecting user data. If you’re buying things online, then you have important behavioral data sets used for modeling. Other kind of target industries would be any industry that uses loyalty cards: airlines, grocery stores. Tracking behavioral data gets you the most interesting, useful, and scientifically valid models.”
Cambridge Analytica worked prominently on both the Ted Cruz and Donald Trump campaigns during the 2016 election season. Asked how the firm made sure the data gathered for each campaign was separate, Kaiser said there were entirely separate staffs and databases for the teams working on the Trump campaign and with the pro-Cruz “Keep the Promise I/Make America Number 1” Super PAC.
Cambridge Analytica also worked on a variety of senatorial and congressional campaigns, as well as state GOP races going back to the 2014 midterm elections. Kaiser reiterated that she didn’t work in operations or directly on campaigns, but gave a few specific examples, including President Trump’s new National Security Advisor.
“In the 2014 midterms, we did Senator Tom Cotton and worked for John Bolton’s Super PAC. Bolton has had a Super PAC for years that he uses that to endorse candidates who are strong on national security policies,” said Kaiser. “This is something all campaigns do. I think people are just outraged over Mr. Trump becoming president and they can’t seem to come to grips with the fact that the Trump Campaign probably was using the exact same tactics as Obama did in 2012.”
Cambridge Pitched Elections All Over the World
When Kaiser started at Cambridge Analytica, the firm was very small, she said. They had one psychologist, a couple data scientists and messaging experts, and a few execs. She and Alexander Nix were the only ones doing sales. By the time she left, Cambridge had its own large-scale digital shop developing its own software and ad technology. But Kaiser maintains that Cambridge is a “boutique shop” compared to data mining companies such as Palantir.
Kaiser was meeting with clients all over the world, often pitching Cambridge’s data analytics capabilities to political candidates. Cambridge’s work on the Nigerian presidential election in 2015 has received a lot of attention, but Kaiser detailed elections she pitched (some of which the company later worked on) all over the world.
“We pitched for work in Germany and France, in Lithuania, in Malaysia, in the Philippines, in Mexico. I ran our Mexico City office for a while, which I originally built for working on commercial projects. I did a lot of political pitching and took some research, but I never ran a campaign—otherwise, I’d be there right now. The [Mexico] election’s on July 1st,” said Kaiser.
“I explored doing work in Colombia but never did the meetings. I know we did political pitches in Argentina, but I don’t think we did the work. I wasn’t involved in Kenya, but our company was. And I did pitch work for Ethiopia, Romania. It just kind of is endless.”
What Happened When the Whistle Blew
Kaiser admits that for most of her time at Cambridge and SCL, she didn’t question the data-gathering practices. When former employee Christopher Wylie blew the whistle earlier this year and turned data and documents over to The Guardian, Kaiser had a long overdue wake-up call.
“When I started getting more involved in the development of blockchain platforms and advising companies in that sector, I started to feel slightly uneasy about what was going on. I didn’t fully question it or challenge it head-on until Chris Wylie came out. To be honest, it’s hard when you’re knee-deep in the trenches of an industry to see it how others see it.”
After the whistleblowing, Kaiser wanted out. She confronted Nix and Cambridge’s leadership and said she didn’t feel safe working in politics anymore without legal guarantees.
“The press kept on accusing us of not obeying regulations, which I kind of wondered if we were or not. I have legal training, but I’m not an expert in every jurisdiction I get sent to,” she said. “They responded poorly at first, and then I invited our former CEO, Alexander Nix, and our current Chief Data Officer, Dr. Alex Taylor, to come to Davos to the World Economic Forum where I co-organized a blockchain meetup called CryptoHQ.
Kaiser had already started consulting for other companies, particularly in the blockchain sector, when Wylie turned the company upside down. After Davos she also began working on blockchain tech for Cambridge Analytica, which was part of now-scrapped plans for an initial coin offering (ICO).
“I started consulting alongside my job with Cambridge Analytica, which Alexander didn’t like very much,” she said. “Looking back now I feel like the company executives probably didn’t have my best interest at heart. The [undercover Channel 4 report] made me question a lot of things, which is why I started looking through old documents and emails, to see if I could find any evidence of wrongdoing.”
She turned that evidence over to US and UK authorities.
Cambridge Analytica obtained data on as many as 87 million Facebook users, but Kaiser said during her UK parliamentary testimony that the true figure is likely much larger. When I asked what she meant by that, Kaiser said it’s not necessarily from Cambridge Analytica but from scores of other companies that have done the same exact thing.
“If we take time to read terms and conditions, our data has been harvested. It’s been collected, modeled, and monetized, sometimes sold as raw data and sometimes licensed for advertisers to target us,” Kaiser said.
“I’ve spent the past four years working professionally in data science-as-a-service. What I’ve always said to governments, to private companies, and to individuals is that no matter what you’re selling, the data you collect is the most valuable thing about your company or organization. Data is the most valuable asset in the world.”
Originally published at www.pcmag.com.