The scandal that is wiping billions of dollars in value off the world’s richest company is about much more than just social media and data mining, writes Michael Brull.
British news program on Channel 4 has exposed Cambridge Analytica and Facebook for what has become an international scandal. However, due to the size and complexity of the issues raised, I don’t think their significance or implications have received the analysis they warrant. In this two part series, I’m going to lay out what I think are the important elements of the story. Firstly, I will show why Cambridge Analytica itself is a big deal. In part 2, I will turn to why we should be worried about other Cambridge Analyticas which have not yet been exposed.
They say they are very big, they say they work around the world, and they say they get results
The series of videos exposing Cambridge Analytica – the lead video has some 2 million views – is dominated by the undercover video recordings of leading figures in the company bragging about their achievements. They are working in America, Africa, Mexico, Malaysia, Australia (!), Brazil, and China – though “not in politics” in China. They said they ran a “very successful project” in an Eastern European country, and “no-one even knew they were there”. They produced “very good material”, they “ghosted in” and “ghosted out”. They bragged about running the campaign of the incumbent Kenyan president in 2013 and 2017. In both cases in Kenya, their campaign was successful.
They also say they won Trump the US Presidential election. As they said, “We did all the research, all the data, all the analytics, all the targeting. We ran all the digital campaign, the digital campaign, the television campaign and our data informed all the strategy”. And despite losing the popular vote by three million, Trump still won, supposedly due to their electoral genius.
One might think this is empty braggadocio. They are saying they are immensely good at what they do. So one might ask: is there evidence of them being effective at campaigning? Did they really win Trump the election?
Cambridge Analytica really did play a central role in electing Trump
In November 2016, Forbes ran a puff piece on Jared Kushner, the son-in-law of Donald Trump, who had just been elected President of the United States. As it was the closest thing to an authorised look at the campaign, it gave good insight into how the central players regarded the campaign, and how the campaign had functioned. It stressed the central role of social media in the campaign.
Titled “How Jared Kushner Won Trump The White House”, it explains Kushner’s central role in running Trump’s campaign. Steven Bertoni reported:
“It’s hard to overstate and hard to summarize Jared’s role in the campaign,” says billionaire Peter Thiel, the only significant Silicon Valley figure to publicly back Trump. “If Trump was the CEO, Jared was effectively the chief operating officer.’” Kushner was involved in the campaign from the early days, going “all in” in November 2015. Kushner basically established an “actual campaign operation”. And he argued to Trump that the campaign was “underutilizing social media. The candidate, in turn, asked his son-in-law to take over his Facebook initiatives.”
“I called somebody who works for one of the technology companies that I work with, and I had them give me a tutorial on how to use Facebook micro-targeting,” Kushner says. They went from selling $8,000 of hats and other Trump merchandise each day, to selling $80,000. They began aggressively testing each Facebook initiative, to get the best return on investment, and maximise reach at cheapest cost. As Bertoni noted, “Television and online advertising? Small and smaller. Twitter and Facebook would fuel the campaign, as key tools for not only spreading Trump’s message but also targeting potential supporters, scraping massive amounts of constituent data and sensing shifts in sentiment in real time.”
Kushner said, “We weren’t afraid to make changes. We weren’t afraid to fail. We tried to do things very cheaply, very quickly. And if it wasn’t working, we would kill it quickly… It meant making quick decisions, fixing things that were broken and scaling things that worked.”
How effective this is, depends on the data they have, how well they are analysing their data, how good they are at reading reaction to news stories and their own messaging, and how good they are at finding the right audience to pitch their messages to. In short, they needed experts to analyse the data for them, on a massive scale. Bertoni noted:
“Kushner’s crew was able to tap into the Republican National Committee’s data machine, and it hired targeting partners like Cambridge Analytica to map voter universes and identify which parts of the Trump platform mattered most: trade, immigration or change.”
“Soon the data operation dictated every campaign decision: travel, fundraising, advertising, rally locations – even the topics of the speeches.”
The data ruled the campaign. “Ineffective ads were killed in minutes, while successful ones scaled. The campaign was sending more than 100,000 uniquely tweaked ads to targeted voters each day.” They raised over $250 million in four months, “mostly from small donors”. Kushner’s system, relying on “up-to-the-minute voter data, provided both ample cash and the insight on where to spend it. When the campaign registered the fact that momentum in Michigan and Pennsylvania was turning Trump’s way, Kushner unleashed tailored TV ads, last-minute rallies and thousands of volunteers to knock on doors and make phone calls.” Those states were crucial for Trump’s victory.
What was the significance of Cambridge Analytica in Kushner’s data-based campaign? As Matt Taibbi, among the most astute observers of Trump’s campaign observed, “Trump on August 17th of [2016] was in the 27th day of an incredible poll tailspin”. Trump was trailing Clinton by double digits in key states, and “There was seemingly no path to victory”. Then, Steve Bannon joined Trump’s team: “As soon as Bannon came aboard, Trump changed course in an unexpected direction. One change was stylistic. He started to read from prepared remarks on the stump more, cutting down on the off-the-cuff rants that had gotten him into so much trouble.” And there were changes in content. “In less than a month,” Bannon had “not only completely reversed Trump’s tailspin, but helped him gain ground besides. By mid-September, polls showed Trump within two points of Hillary, or even in a dead heat in some cases.”
Why does this matter? As Vox explains, “Cambridge Analytica was created when Steve Bannon approached conservative megadonors Rebekah and Robert Mercer to fund a political consulting firm.” Bannon became Vice-President of Cambridge Analytica. In June 2016, Cambridge Analytica joined the Trump campaign, and in August Bannon came aboard. The cause and effect perhaps wasn’t immediate, but this is understandable given the nature of the enterprise.
Brittany Kaiser, a former business development director for Cambridge Analytica who turned whistle-blower, said that they had needed to build data from scratch when they joined Trump’s campaign. She said, “There was no database of record. There were many disparate data sources that were not connected, matched or hygiene… There was no data science programme, so they weren’t undertaking any modelling. There was no digital marketing team.”
Whilst Kushner bragged about all of the micro-targeting after victory, it seemed he didn’t know what he was really doing until he received high-powered expert help. Unsurprisingly, it wasn’t quite nepotism that won the campaign for Trump. It was billionaire megadonors funding a sophisticated data analytics team.
That team was able to utilise enormous data on Facebook users and micro-target their campaigns, because of three key features of Facebook. One is that it is extremely popular. Some 68 percent of Americans use it. 74 percent of them use it daily. It is an interactive medium with an enormous audience, which can be targeted and reached.
The second key feature is the extraordinarily detailed feedback you get about everything you post on it. If I post a video on my public page, I will get data on how many people watch it, how long people tend to watch it for and so on. I can use that data to judge how effective my messaging is, and tailor it accordingly. With a large budget, I could experiment, and get a detailed breakdown of what does and doesn’t work.
The other important aspect is Facebook’s contempt for user privacy. When Mark Zuckerberg originally launched Facebook, a friend asked how he was able to get private information from 4,000 people. His responses ran as follows: “I don’t know why”, “They ‘trust me’”, “Dumb fucks”.
Facebook is free, because it offers great value to people who want to sell ads on it. That is, it is a business where the consumers are advertisers, and the product is the audience. Ads can be particularly targeted because of Facebook’s lax approach to user privacy. As people share so much of their lives on Facebook – we post our opinions, share videos and photos of intimate social gatherings, discuss work and TV, message friends, co-workers, lovers and family – all of that information is sucked up by Facebook, and used to target ads.
The fact that it can be used politically should be obvious, but is only one feature of the site. Facebook is a handy site for the public. It is a lucrative business because it is so invasive, which is why Zuckerberg is the fifth richest person on the planet, worth some $72 billion (hopefully plummeting at the time of writing).
Apparently, Cambridge Analytica got data on 50 million Americans through shady means. Basically, some 320,000 people were paid to do a personality quiz. To do it, they logged onto Facebook. Cambridge was then able to harvest their Facebook data, and that of their friends too. Theoretically, they told Facebook this was for academic purposes, but obviously it wouldn’t have been possible if Facebook protected the data of users. Furthermore, they made virtually no effort to protect the data, once they knew it had been illicitly obtained. And why should they? Facebook has never protected user’s data. Their business thrives precisely because of their contempt for the privacy of users.
Cambridge Analytica was not above using illicitly obtained data on over 50 million Americans. They were caught on film by Channel 4 saying they would be comfortable doing things like sending women around to seduce politicians, or otherwise film and expose corruption among opposing politicians. To such a company, what is a bit of stolen data?
With the enormous amount of material they harvested from Facebook, and the extremely detailed feedback on each post, meme and video, the Trump campaign could carefully target each message, fine tuning the appeal each time. The messaging got smoother, and they learned to hit the right notes. That’s how Trump won.
It is hard to imagine an explanation for Trump’s campaign winning, which does not give pride of place to Cambridge Analytica. His campaign significantly underspent the Clinton campaign, particularly on traditional media, and the unofficial leader of the campaign stressed how central social media work was to their election victory. The formal chief executive of the Trump campaign, Steve Bannon, who effectively turned the campaign around, was the guy who set up Cambridge Analytica in the first place, and moved to Trump’s campaign after working there as its vice-president.
So, we don’t know if Cambridge Analytica has achieved what they said they did in the unnamed Eastern European country. We don’t know what kind of effect they’ve had in Africa, Mexico, Malaysia, and we certainly don’t know what they have done in Australia. But when they say they won the election for Trump, they weren’t exactly lying.
The fact that that claim deserved to be taken seriously suggests that we should also be concerned about whether they’ve played a similarly important role in other parts of the world, including places they didn’t feel comfortable bragging about.